Generate faces for yolov3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.596268Z",
     "start_time": "2019-12-30T02:59:57.132145Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Anti\\Anaconda3\\envs\\ml\\lib\\site-packages\\h5py\\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import json\n",
    "import imageio\n",
    "import matplotlib.pyplot as plt\n",
    "import cv2\n",
    "import numpy as np\n",
    "import sklearn\n",
    "import pickle\n",
    "from sklearn.cluster import KMeans\n",
    "from mtcnn.mtcnn import MTCNN\n",
    "from PIL import Image\n",
    "from imageio import imwrite"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Validate face feature detection speed\n",
    "* Train on bigger boxes as it is important to get the face inside the box, box does not have to be that tight"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.663269Z",
     "start_time": "2019-12-30T03:00:05.598246Z"
    },
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "# Video files\n",
    "videos_folder = '../deepfake-detection-challenge/train_sample_videos/'\n",
    "videos = list(set(os.listdir(videos_folder)) - set(['metadata.json']))\n",
    "with open(os.path.join(videos_folder, 'metadata.json')) as json_file:\n",
    "    metadata = json.load(json_file)\n",
    "    \n",
    "# Video into df\n",
    "df = pd.DataFrame(metadata).T\n",
    "df.reset_index(level=0, inplace=True)\n",
    "df.rename(index={'index': 'vide'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.672250Z",
     "start_time": "2019-12-30T03:00:05.664245Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['bmjmjmbglm.mp4',\n",
       " 'bqdjzqhcft.mp4',\n",
       " 'aslsvlvpth.mp4',\n",
       " 'dsgpbgsrdm.mp4',\n",
       " 'bddjdhzfze.mp4',\n",
       " 'dfbpceeaox.mp4',\n",
       " 'avssvvsdhz.mp4',\n",
       " 'btugrnoton.mp4',\n",
       " 'cizlkenljw.mp4',\n",
       " 'crezycjqyk.mp4',\n",
       " 'aorjvbyxhw.mp4',\n",
       " 'atkdltyyen.mp4',\n",
       " 'errocgcham.mp4',\n",
       " 'ellavthztb.mp4',\n",
       " 'acqfdwsrhi.mp4',\n",
       " 'chviwxsfhg.mp4',\n",
       " 'augtsuxpzc.mp4',\n",
       " 'dgzklxjmix.mp4',\n",
       " 'dwediigjit.mp4',\n",
       " 'eprybmbpba.mp4',\n",
       " 'cnilkgvfei.mp4',\n",
       " 'bctvsmddgq.mp4',\n",
       " 'bpapbctoao.mp4',\n",
       " 'eekozbeafq.mp4',\n",
       " 'abofeumbvv.mp4',\n",
       " 'coadfnerlk.mp4',\n",
       " 'awnwkrqibf.mp4',\n",
       " 'dhkwmjxwrn.mp4',\n",
       " 'bggsurpgpr.mp4',\n",
       " 'bffwsjxghk.mp4',\n",
       " 'bbhtdfuqxq.mp4',\n",
       " 'dkzvdrzcnr.mp4',\n",
       " 'btmsngnqhv.mp4',\n",
       " 'bopqhhalml.mp4',\n",
       " 'ajwpjhrbcv.mp4',\n",
       " 'bnbuonyoje.mp4',\n",
       " 'dhevettufk.mp4',\n",
       " 'altziddtxi.mp4',\n",
       " 'dsjbknkujw.mp4',\n",
       " 'eukvucdetx.mp4',\n",
       " 'cycacemkmt.mp4',\n",
       " 'dkrvorliqc.mp4',\n",
       " 'edyncaijwx.mp4',\n",
       " 'bulkxhhknf.mp4',\n",
       " 'emgjphonqb.mp4',\n",
       " 'bgaogsjehq.mp4',\n",
       " 'eczrseixwq.mp4',\n",
       " 'avfitoutyn.mp4',\n",
       " 'btjwbtsgln.mp4',\n",
       " 'amaivqofda.mp4',\n",
       " 'dvakowbgbt.mp4',\n",
       " 'cglxirfaey.mp4',\n",
       " 'eebrkicpry.mp4',\n",
       " 'dhoqofwoxa.mp4',\n",
       " 'ebywfrmhtd.mp4',\n",
       " 'atxvxouljq.mp4',\n",
       " 'avmjormvsx.mp4',\n",
       " 'ekcrtigpab.mp4',\n",
       " 'ahqqqilsxt.mp4',\n",
       " 'ctzmavwror.mp4',\n",
       " 'chtapglbcj.mp4',\n",
       " 'ehccixxzoe.mp4',\n",
       " 'epymyyiblu.mp4',\n",
       " 'cpjxareypw.mp4',\n",
       " 'bguwlyazau.mp4',\n",
       " 'bqeiblbxtl.mp4',\n",
       " 'bvzjkezkms.mp4',\n",
       " 'asaxgevnnp.mp4',\n",
       " 'bsqgziaylx.mp4',\n",
       " 'acxwigylke.mp4',\n",
       " 'crzfebnfgb.mp4',\n",
       " 'ecuvtoltue.mp4',\n",
       " 'eivxffliio.mp4',\n",
       " 'dbhoxkblzx.mp4',\n",
       " 'dtbpmdqvao.mp4',\n",
       " 'djvutyvaio.mp4',\n",
       " 'ehdkmxgtxh.mp4',\n",
       " 'bgwmmujlmc.mp4',\n",
       " 'bmehkyanbj.mp4',\n",
       " 'elginszwtk.mp4',\n",
       " 'bofqajtwve.mp4',\n",
       " 'ciyoudyhly.mp4',\n",
       " 'clihsshdkq.mp4',\n",
       " 'ckkuyewywx.mp4',\n",
       " 'bmioepcpsx.mp4',\n",
       " 'esyrimvzsa.mp4',\n",
       " 'bilnggbxgu.mp4',\n",
       " 'bkmdzhfzfh.mp4',\n",
       " 'bqkdbcqjvb.mp4',\n",
       " 'duycddgtrl.mp4',\n",
       " 'cferslmfwh.mp4',\n",
       " 'cksanfsjhc.mp4',\n",
       " 'bxzakyopjf.mp4',\n",
       " 'cqrskwiqng.mp4',\n",
       " 'bjsmaqefoi.mp4',\n",
       " 'eepezmygaq.mp4',\n",
       " 'eoewqcpbgt.mp4',\n",
       " 'egbbcxcuqy.mp4',\n",
       " 'etejaapnxh.mp4',\n",
       " 'acxnxvbsxk.mp4',\n",
       " 'arlmiizoob.mp4',\n",
       " 'crktehraph.mp4',\n",
       " 'dptrzdvwpg.mp4',\n",
       " 'dntkzzzcdh.mp4',\n",
       " 'cffffbcywc.mp4',\n",
       " 'bdxuhamuqx.mp4',\n",
       " 'afoovlsmtx.mp4',\n",
       " 'ekhacizpah.mp4',\n",
       " 'avvdgsennp.mp4',\n",
       " 'andaxzscny.mp4',\n",
       " 'cxttmymlbn.mp4',\n",
       " 'cgvrgibpfo.mp4',\n",
       " 'cwsbspfzck.mp4',\n",
       " 'bgvhtpzknn.mp4',\n",
       " 'dbzcqmxzaj.mp4',\n",
       " 'axczxisdtb.mp4',\n",
       " 'bhbdugnurr.mp4',\n",
       " 'dkwjwbwgey.mp4',\n",
       " 'djxdyjopjd.mp4',\n",
       " 'bpxckdzddv.mp4',\n",
       " 'cvaksbpssm.mp4',\n",
       " 'diopzaywor.mp4',\n",
       " 'bdbhekrrwo.mp4',\n",
       " 'dqqtjcryjv.mp4',\n",
       " 'ehieahnhte.mp4',\n",
       " 'awhmfnnjih.mp4',\n",
       " 'blzydqdfem.mp4',\n",
       " 'diomeixhrg.mp4',\n",
       " 'byyqectxqa.mp4',\n",
       " 'axwgcsyphv.mp4',\n",
       " 'dubiroskqn.mp4',\n",
       " 'ddjggcasdw.mp4',\n",
       " 'dhcselezer.mp4',\n",
       " 'dsndhujjjb.mp4',\n",
       " 'bpwzipqtxf.mp4',\n",
       " 'dafhtipaml.mp4',\n",
       " 'duvyaxbzvp.mp4',\n",
       " 'aczrgyricp.mp4',\n",
       " 'cfxkpiweqt.mp4',\n",
       " 'dzvyfiarrq.mp4',\n",
       " 'asdpeebotb.mp4',\n",
       " 'cdyakrxkia.mp4',\n",
       " 'asvcrfdpnq.mp4',\n",
       " 'brhalypwoo.mp4',\n",
       " 'dbhrpizyeq.mp4',\n",
       " 'ckjaibzfxa.mp4',\n",
       " 'ddpvuimigj.mp4',\n",
       " 'aagfhgtpmv.mp4',\n",
       " 'cepxysienc.mp4',\n",
       " 'aettqgevhz.mp4',\n",
       " 'czmqpxrqoh.mp4',\n",
       " 'bqtuuwzdtr.mp4',\n",
       " 'ahbweevwpv.mp4',\n",
       " 'dlpoieqvfb.mp4',\n",
       " 'caqbrkogkb.mp4',\n",
       " 'dhcndnuwta.mp4',\n",
       " 'dgmevclvzy.mp4',\n",
       " 'bwhlgysghg.mp4',\n",
       " 'efwfxwwlbw.mp4',\n",
       " 'aybgughjxh.mp4',\n",
       " 'atzdznmder.mp4',\n",
       " 'dtocdfbwca.mp4',\n",
       " 'aqpnvjhuzw.mp4',\n",
       " 'dakqwktlbi.mp4',\n",
       " 'curpwogllm.mp4',\n",
       " 'bnjcdrfuov.mp4',\n",
       " 'bmjzrlszhi.mp4',\n",
       " 'eahlqmfvtj.mp4',\n",
       " 'cdbsbdymzd.mp4',\n",
       " 'cyclgfjdrv.mp4',\n",
       " 'dakiztgtnw.mp4',\n",
       " 'efdyrflcpg.mp4',\n",
       " 'axoygtekut.mp4',\n",
       " 'bsfmwclnqy.mp4',\n",
       " 'bhsluedavd.mp4',\n",
       " 'byijojkdba.mp4',\n",
       " 'apatcsqejh.mp4',\n",
       " 'ebebgmtlcu.mp4',\n",
       " 'ayqvfdhslr.mp4',\n",
       " 'aladcziidp.mp4',\n",
       " 'cwwandrkus.mp4',\n",
       " 'degpbqvcay.mp4',\n",
       " 'bhaaboftbc.mp4',\n",
       " 'eggbjzxnmg.mp4',\n",
       " 'dkhlttuvmx.mp4',\n",
       " 'drtbksnpol.mp4',\n",
       " 'emaalmsonj.mp4',\n",
       " 'adhsbajydo.mp4',\n",
       " 'azpuxunqyo.mp4',\n",
       " 'bseamdrpbj.mp4',\n",
       " 'cprhtltsjp.mp4',\n",
       " 'byunigvnay.mp4',\n",
       " 'dcamvmuors.mp4',\n",
       " 'ahdbuwqxit.mp4',\n",
       " 'beyebyhrph.mp4',\n",
       " 'egghxjjmfg.mp4',\n",
       " 'eiwopxzjfn.mp4',\n",
       " 'apogckdfrz.mp4',\n",
       " 'btohlidmru.mp4',\n",
       " 'cdaxixbosp.mp4',\n",
       " 'doanjploai.mp4',\n",
       " 'beboztfcme.mp4',\n",
       " 'amowujxmzc.mp4',\n",
       " 'bdgipnyobr.mp4',\n",
       " 'czkdanyadc.mp4',\n",
       " 'ecwaxgutkc.mp4',\n",
       " 'bwipwzzxxu.mp4',\n",
       " 'bqhtpqmmqp.mp4',\n",
       " 'dnyvfblxpm.mp4',\n",
       " 'aelfnikyqj.mp4',\n",
       " 'etohcvnzbj.mp4',\n",
       " 'aneclqfpbt.mp4',\n",
       " 'aipfdnwpoo.mp4',\n",
       " 'adohikbdaz.mp4',\n",
       " 'diqraixiov.mp4',\n",
       " 'ensyyivobf.mp4',\n",
       " 'etmcruaihe.mp4',\n",
       " 'awukslzjra.mp4',\n",
       " 'dkuayagnmc.mp4',\n",
       " 'ejkqesyvam.mp4',\n",
       " 'brwrlczjvi.mp4',\n",
       " 'apgjqzkoma.mp4',\n",
       " 'drgjzlxzxj.mp4',\n",
       " 'btiysiskpf.mp4',\n",
       " 'aevrfsexku.mp4',\n",
       " 'arrhsnjqku.mp4',\n",
       " 'ddhfabwpuz.mp4',\n",
       " 'asmpfjfzif.mp4',\n",
       " 'dptbnjnkdg.mp4',\n",
       " 'bbhpvrmbse.mp4',\n",
       " 'eeyhxisdfh.mp4',\n",
       " 'cyboodqqyr.mp4',\n",
       " 'dvumqqhoac.mp4',\n",
       " 'eckvhdusax.mp4',\n",
       " 'cdphtzqrvp.mp4',\n",
       " 'dzqwgqewhu.mp4',\n",
       " 'bghphrsfxf.mp4',\n",
       " 'avnqydkqjj.mp4',\n",
       " 'ebeknhudxq.mp4',\n",
       " 'byofowlkki.mp4',\n",
       " 'cobjrlugvp.mp4',\n",
       " 'esckbnkkvb.mp4',\n",
       " 'byqzyxifza.mp4',\n",
       " 'dhxctgyoqj.mp4',\n",
       " 'bkwxhglwct.mp4',\n",
       " 'cwrtyzndpx.mp4',\n",
       " 'bmhvktyiwp.mp4',\n",
       " 'cxfujlvsuw.mp4',\n",
       " 'avywawptfc.mp4',\n",
       " 'esxrvsgpvb.mp4',\n",
       " 'dgxrqjdomn.mp4',\n",
       " 'dozyddhild.mp4',\n",
       " 'azsmewqghg.mp4',\n",
       " 'btunxncpjh.mp4',\n",
       " 'dsdoseflas.mp4',\n",
       " 'czfunozvwp.mp4',\n",
       " 'ehevsxtecd.mp4',\n",
       " 'dboxtiehng.mp4',\n",
       " 'boovltmuwi.mp4',\n",
       " 'ceymbecxnj.mp4',\n",
       " 'axntxmycwd.mp4',\n",
       " 'aufmsmnoye.mp4',\n",
       " 'bndybcqhfr.mp4',\n",
       " 'ctpqeykqdp.mp4',\n",
       " 'cmxcfkrjiv.mp4',\n",
       " 'ckbdwedgmc.mp4',\n",
       " 'cbbibzcoih.mp4',\n",
       " 'bydaidkpdp.mp4',\n",
       " 'ekkdjkirzq.mp4',\n",
       " 'btjlfpzbdu.mp4',\n",
       " 'bweezhfpzp.mp4',\n",
       " 'aelzhcnwgf.mp4',\n",
       " 'eudeqjhdfd.mp4',\n",
       " 'bhpwpydzpo.mp4',\n",
       " 'bqqpbzjgup.mp4',\n",
       " 'dxuliowugt.mp4',\n",
       " 'ehbnclaukr.mp4',\n",
       " 'bjjbwsqjir.mp4',\n",
       " 'cknyxaqouy.mp4',\n",
       " 'anpuvshzoo.mp4',\n",
       " 'bvgwelbeof.mp4',\n",
       " 'bzmdrafeex.mp4',\n",
       " 'drcyabprvt.mp4',\n",
       " 'esgftaficx.mp4',\n",
       " 'ebchwmwayp.mp4',\n",
       " 'eqvuznuwsa.mp4',\n",
       " 'esnntzzajv.mp4',\n",
       " 'cmbzllswnl.mp4',\n",
       " 'aapnvogymq.mp4',\n",
       " 'etdcqxabww.mp4',\n",
       " 'bzythlfnhq.mp4',\n",
       " 'ecujsjhscd.mp4',\n",
       " 'cqfugiqupm.mp4',\n",
       " 'dxuplhwvig.mp4',\n",
       " 'aklqzsddfl.mp4',\n",
       " 'emfbhytfhc.mp4',\n",
       " 'djvtbgwdcc.mp4',\n",
       " 'bjkmjilrxp.mp4',\n",
       " 'bchnbulevv.mp4',\n",
       " 'dkdwxmtpuo.mp4',\n",
       " 'alaijyygdv.mp4',\n",
       " 'bdnaqemxmr.mp4',\n",
       " 'akxoopqjqz.mp4',\n",
       " 'byfenovjnf.mp4',\n",
       " 'dbnygxtwek.mp4',\n",
       " 'dbzpcjntve.mp4',\n",
       " 'dzyuwjkjui.mp4',\n",
       " 'axwovszumc.mp4',\n",
       " 'bkvetcojbt.mp4',\n",
       " 'cbltdtxglo.mp4',\n",
       " 'ccmonzqfrz.mp4',\n",
       " 'erqgqacbqe.mp4',\n",
       " 'cwbacdwrzo.mp4',\n",
       " 'btxlttbpkj.mp4',\n",
       " 'dofusvhnib.mp4',\n",
       " 'cqhngvpgyi.mp4',\n",
       " 'elvvackpjh.mp4',\n",
       " 'cuzrgrbvil.mp4',\n",
       " 'cyxlcuyznd.mp4',\n",
       " 'erlvuvjsjf.mp4',\n",
       " 'bbvgxeczei.mp4',\n",
       " 'covdcysmbi.mp4',\n",
       " 'deywhkarol.mp4',\n",
       " 'cthdnahrkh.mp4',\n",
       " 'cttqtsjvgn.mp4',\n",
       " 'bqnymlsayl.mp4',\n",
       " 'eixwxvxbbn.mp4',\n",
       " 'atyntldecu.mp4',\n",
       " 'duzuusuajr.mp4',\n",
       " 'eejswgycjc.mp4',\n",
       " 'esyhwdfnxs.mp4',\n",
       " 'benmsfzfaz.mp4',\n",
       " 'adylbeequz.mp4',\n",
       " 'chzieimrwu.mp4',\n",
       " 'avtycwsgyb.mp4',\n",
       " 'dnhvalzvrt.mp4',\n",
       " 'avibnnhwhp.mp4',\n",
       " 'dxbqjxrhin.mp4',\n",
       " 'cfyduhpbps.mp4',\n",
       " 'blpchvmhxx.mp4',\n",
       " 'cettndmvzl.mp4',\n",
       " 'aybumesmpk.mp4',\n",
       " 'dqnyszdong.mp4',\n",
       " 'diuzrpqjli.mp4',\n",
       " 'ehtdtkmmli.mp4',\n",
       " 'abqwwspghj.mp4',\n",
       " 'dqzreruvje.mp4',\n",
       " 'aknbdpmgua.mp4',\n",
       " 'bwuwstvsbw.mp4',\n",
       " 'akzbnazxtz.mp4',\n",
       " 'cwqlvzefpg.mp4',\n",
       " 'bmbbkwmxqj.mp4',\n",
       " 'aytzyidmgs.mp4',\n",
       " 'eajlrktemq.mp4',\n",
       " 'arkroixhey.mp4',\n",
       " 'clrycekyst.mp4',\n",
       " 'acifjvzvpm.mp4',\n",
       " 'eqnoqyfquo.mp4',\n",
       " 'eebserckhh.mp4',\n",
       " 'dcuiiorugd.mp4',\n",
       " 'dlrsbscitn.mp4',\n",
       " 'dulanfulol.mp4',\n",
       " 'dzwkmcwkwl.mp4',\n",
       " 'dqswpjoepo.mp4',\n",
       " 'alvgwypubw.mp4',\n",
       " 'ddepeddixj.mp4',\n",
       " 'agrmhtjdlk.mp4',\n",
       " 'eqjscdagiv.mp4',\n",
       " 'ajqslcypsw.mp4',\n",
       " 'cxrfacemmq.mp4',\n",
       " 'drsakwyvqv.mp4',\n",
       " 'bejhvclboh.mp4',\n",
       " 'ddqccgmtka.mp4',\n",
       " 'ahfazfbntc.mp4',\n",
       " 'dhjmzhrcav.mp4',\n",
       " 'agqphdxmwt.mp4',\n",
       " 'bahdpoesir.mp4',\n",
       " 'caifxvsozs.mp4',\n",
       " 'bntlodcfeg.mp4',\n",
       " 'dbtbbhakdv.mp4',\n",
       " 'ccfoszqabv.mp4',\n",
       " 'avgiuextiz.mp4',\n",
       " 'aknmpoonls.mp4',\n",
       " 'deyyistcrd.mp4',\n",
       " 'dnexlwbcxq.mp4',\n",
       " 'cppdvdejkc.mp4',\n",
       " 'eiriyukqqy.mp4',\n",
       " 'abarnvbtwb.mp4',\n",
       " 'dqppxmoqdl.mp4',\n",
       " 'atvmxvwyns.mp4',\n",
       " 'alninxcyhg.mp4',\n",
       " 'ehfiekigla.mp4',\n",
       " 'agdkmztvby.mp4',\n",
       " 'brvqtabyxj.mp4',\n",
       " 'bgmlwsoamc.mp4',\n",
       " 'dzieklokdr.mp4',\n",
       " 'bourlmzsio.mp4',\n",
       " 'akvmwkdyuv.mp4',\n",
       " 'ecnihjlfyt.mp4',\n",
       " 'ebkzwjgjhq.mp4']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "videos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.687272Z",
     "start_time": "2019-12-30T03:00:05.673247Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>label</th>\n",
       "      <th>split</th>\n",
       "      <th>original</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>aagfhgtpmv.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>vudstovrck.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>aapnvogymq.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>jdubbvfswz.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>abarnvbtwb.mp4</td>\n",
       "      <td>REAL</td>\n",
       "      <td>train</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>abofeumbvv.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>atvmxvwyns.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>abqwwspghj.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>qzimuostzz.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>etejaapnxh.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>wtreibcmgm.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>396</th>\n",
       "      <td>etmcruaihe.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>afoovlsmtx.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>397</th>\n",
       "      <td>etohcvnzbj.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>bdnaqemxmr.mp4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>398</th>\n",
       "      <td>eudeqjhdfd.mp4</td>\n",
       "      <td>REAL</td>\n",
       "      <td>train</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399</th>\n",
       "      <td>eukvucdetx.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>gjypopglvi.mp4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>400 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              index label  split        original\n",
       "0    aagfhgtpmv.mp4  FAKE  train  vudstovrck.mp4\n",
       "1    aapnvogymq.mp4  FAKE  train  jdubbvfswz.mp4\n",
       "2    abarnvbtwb.mp4  REAL  train            None\n",
       "3    abofeumbvv.mp4  FAKE  train  atvmxvwyns.mp4\n",
       "4    abqwwspghj.mp4  FAKE  train  qzimuostzz.mp4\n",
       "..              ...   ...    ...             ...\n",
       "395  etejaapnxh.mp4  FAKE  train  wtreibcmgm.mp4\n",
       "396  etmcruaihe.mp4  FAKE  train  afoovlsmtx.mp4\n",
       "397  etohcvnzbj.mp4  FAKE  train  bdnaqemxmr.mp4\n",
       "398  eudeqjhdfd.mp4  REAL  train            None\n",
       "399  eukvucdetx.mp4  FAKE  train  gjypopglvi.mp4\n",
       "\n",
       "[400 rows x 4 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load video"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.694245Z",
     "start_time": "2019-12-30T03:00:05.688248Z"
    }
   },
   "outputs": [],
   "source": [
    "class Video:\n",
    "    def __init__(self, path):\n",
    "        self.path = path\n",
    "        self.container = imageio.get_reader(path, 'ffmpeg')\n",
    "        self.length = self.container.count_frames()\n",
    "        self.fps = self.container.get_meta_data()['fps']\n",
    "    \n",
    "    def init_head(self):\n",
    "        self.container.set_image_index(0)\n",
    "    \n",
    "    def next_frame(self):\n",
    "        self.container.get_next_data()\n",
    "    \n",
    "    def get(self, key):\n",
    "        return self.container.get_data(key)\n",
    "    \n",
    "    def __call__(self, key):\n",
    "        return self.get(key)\n",
    "    \n",
    "    def __len__(self):\n",
    "        return self.length"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare faces for SSD300/SSD512"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.699245Z",
     "start_time": "2019-12-30T03:00:05.695246Z"
    }
   },
   "outputs": [],
   "source": [
    "def save_obj(obj, name, path):\n",
    "    with open(path + name + '.pkl', 'wb') as f:\n",
    "        pickle.dump(obj, f, pickle.HIGHEST_PROTOCOL)\n",
    "\n",
    "def load_obj(name, path):\n",
    "    with open(path + name + '.pkl', 'rb') as f:\n",
    "        return pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:05.710244Z",
     "start_time": "2019-12-30T03:00:05.700246Z"
    }
   },
   "outputs": [],
   "source": [
    "class FacesFromVideo:\n",
    "    def __init__(self, video, video_file_name, frame_interval, detector, total_frames=None):\n",
    "        self.video = video\n",
    "        self.detector = detector\n",
    "        self.video_file_name = video_file_name\n",
    "        self.frame_interval = frame_interval\n",
    "        self.max_frame = video.__len__() if total_frames == None \\\n",
    "            else np.minimum(total_frames * self.frame_interval, video.__len__())\n",
    "        \n",
    "    def extract_faces_from_img(self, img, frame_nr):\n",
    "        faces = {}\n",
    "        results = self.detector.detect_faces(img)\n",
    "        for idx, res in enumerate(results): # Iterate each face per image\n",
    "            box = np.array(res['box'])\n",
    "            box[box < 0] = 0\n",
    "            x1, y1, width, height = box\n",
    "            x2, y2 = x1 + width, y1 + height\n",
    "            face = img[y1:y2, x1:x2]\n",
    "            key = str(str(frame_nr) + '_' + str(idx))\n",
    "            if res['confidence'] > 0.90:\n",
    "                faces[key] = {}\n",
    "                faces[key]['coordinates'] = {'x1': x1, \n",
    "                                             'y1': y1, \n",
    "                                             'width': width, \n",
    "                                             'height': height}\n",
    "                faces[key]['keypoints'] = res['keypoints']\n",
    "                faces[key]['confidence'] = res['confidence']\n",
    "\n",
    "        return faces\n",
    "    \n",
    "    def save_img(self, img, frame_nr):\n",
    "        image = Image.fromarray(img)\n",
    "        image.save('../face_detection/{}_{}.{}'.format(self.video_file_name.split('.')[0], frame_nr, 'png'))\n",
    "    \n",
    "    def start(self):\n",
    "        faces_per_frame = {}\n",
    "        for frame_nr in range(0, self.max_frame, self.frame_interval):\n",
    "            img = video.get(frame_nr)\n",
    "            try:\n",
    "                faces = self.extract_faces_from_img(img, frame_nr)\n",
    "                self.save_img(img, frame_nr)\n",
    "                faces_per_frame.update(faces)\n",
    "            except KeyboardInterrupt:\n",
    "                print('Interrupted')\n",
    "                raise\n",
    "            except:\n",
    "                print('something went wrong with ' + self.video_file_name)\n",
    "\n",
    "        return faces_per_frame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:07.194247Z",
     "start_time": "2019-12-30T03:00:05.712245Z"
    }
   },
   "outputs": [],
   "source": [
    "train_path = '../deepfake_train_full'\n",
    "train_folders = [os.path.join(train_path, x) for x in os.listdir(train_path) if x[-1] != 'p']\n",
    "train_video_folders = [os.path.join(x, os.listdir(x)[0]) for x in train_folders]\n",
    "detector = MTCNN()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T03:00:07.403268Z",
     "start_time": "2019-12-30T03:00:07.195245Z"
    }
   },
   "outputs": [],
   "source": [
    "face_coordinates_per_video = load_obj('face_coordinates_per_video', '../face_detection/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T12:48:45.108015Z",
     "start_time": "2019-12-30T03:00:07.404248Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-10-ac581eed6896>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      7\u001b[0m         faces_from_video = FacesFromVideo(video=video, video_file_name=video_file_name, \\\n\u001b[0;32m      8\u001b[0m                                           frame_interval=75, detector=detector, total_frames=3)\n\u001b[1;32m----> 9\u001b[1;33m         \u001b[0mfaces_per_frame\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mfaces_from_video\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstart\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     10\u001b[0m         \u001b[0mface_coordinates_per_video\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mvideo_file_name\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mfaces_per_frame\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     11\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0midx\u001b[0m \u001b[1;33m%\u001b[0m \u001b[1;36m20\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-7-d32ddd7e66c9>\u001b[0m in \u001b[0;36mstart\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m     36\u001b[0m         \u001b[0mfaces_per_frame\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     37\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mframe_nr\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmax_frame\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mframe_interval\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 38\u001b[1;33m             \u001b[0mimg\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mvideo\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe_nr\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     39\u001b[0m             \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     40\u001b[0m                 \u001b[0mfaces\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mextract_faces_from_img\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mframe_nr\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-5-0175b86cd53d>\u001b[0m in \u001b[0;36mget\u001b[1;34m(self, key)\u001b[0m\n\u001b[0;32m     13\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 15\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontainer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_data\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     16\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     17\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\core\\format.py\u001b[0m in \u001b[0;36mget_data\u001b[1;34m(self, index, **kwargs)\u001b[0m\n\u001b[0;32m    337\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_BaseReaderWriter_last_index\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mindex\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    338\u001b[0m             \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 339\u001b[1;33m                 \u001b[0mim\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmeta\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_get_data\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mindex\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    340\u001b[0m             \u001b[1;32mexcept\u001b[0m \u001b[0mStopIteration\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    341\u001b[0m                 \u001b[1;32mraise\u001b[0m \u001b[0mIndexError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mindex\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\plugins\\ffmpeg.py\u001b[0m in \u001b[0;36m_get_data\u001b[1;34m(self, index)\u001b[0m\n\u001b[0;32m    381\u001b[0m                     \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_initialize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mindex\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    382\u001b[0m                 \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 383\u001b[1;33m                     \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_skip_frames\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mindex\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_pos\u001b[0m \u001b[1;33m-\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    384\u001b[0m                 \u001b[0mresult\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mis_new\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_read_frame\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    385\u001b[0m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_pos\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mindex\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\plugins\\ffmpeg.py\u001b[0m in \u001b[0;36m_skip_frames\u001b[1;34m(self, n)\u001b[0m\n\u001b[0;32m    471\u001b[0m             \u001b[1;34m\"\"\" Reads and throws away n frames \"\"\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    472\u001b[0m             \u001b[1;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mn\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 473\u001b[1;33m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_read_gen\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__next__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    474\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_pos\u001b[0m \u001b[1;33m+=\u001b[0m \u001b[0mn\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    475\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio_ffmpeg\\_io.py\u001b[0m in \u001b[0;36mread_frames\u001b[1;34m(path, pix_fmt, bpp, input_params, output_params)\u001b[0m\n\u001b[0;32m    166\u001b[0m                 \u001b[0mbb\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mbytes\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    167\u001b[0m                 \u001b[1;32mwhile\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbb\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m<\u001b[0m \u001b[0mframesize\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 168\u001b[1;33m                     \u001b[0mextra_bytes\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstdout\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframesize\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbb\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    169\u001b[0m                     \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mextra_bytes\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    170\u001b[0m                         \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbb\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "# face_coordinates_per_video = {}\n",
    "for train_video_folder in train_video_folders[9:]:\n",
    "    train_videos = [os.path.join(train_video_folder, x) for x in os.listdir(train_video_folder) if x[-1] == '4']\n",
    "    for idx, train_video in enumerate(train_videos):\n",
    "        video_file_name = train_video.split('\\\\')[-1]\n",
    "        video = Video(train_video)\n",
    "        faces_from_video = FacesFromVideo(video=video, video_file_name=video_file_name, \\\n",
    "                                          frame_interval=75, detector=detector, total_frames=3)\n",
    "        faces_per_frame = faces_from_video.start()\n",
    "        face_coordinates_per_video[video_file_name] = faces_per_frame\n",
    "        if idx % 20 == 0:\n",
    "            save_obj(face_coordinates_per_video, 'face_coordinates_per_video', '../face_detection/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T12:49:02.085445Z",
     "start_time": "2019-12-30T12:49:02.080469Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'../deepfake_train_full\\\\dfdc_train_part_09\\\\dfdc_train_part_9'"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_video_folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-30T12:49:11.618004Z",
     "start_time": "2019-12-30T12:49:11.613997Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['../deepfake_train_full\\\\dfdc_train_part_09\\\\dfdc_train_part_9',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_10\\\\dfdc_train_part_10',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_11\\\\dfdc_train_part_11',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_12\\\\dfdc_train_part_12',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_13\\\\dfdc_train_part_13',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_14\\\\dfdc_train_part_14',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_15\\\\dfdc_train_part_15',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_16\\\\dfdc_train_part_16',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_17\\\\dfdc_train_part_17',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_18\\\\dfdc_train_part_18',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_19\\\\dfdc_train_part_19',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_20\\\\dfdc_train_part_20',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_21\\\\dfdc_train_part_21',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_22\\\\dfdc_train_part_22',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_23\\\\dfdc_train_part_23',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_24\\\\dfdc_train_part_24',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_25\\\\dfdc_train_part_25',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_26\\\\dfdc_train_part_26',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_27\\\\dfdc_train_part_27',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_28\\\\dfdc_train_part_28',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_29\\\\dfdc_train_part_29',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_30\\\\dfdc_train_part_30',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_31\\\\dfdc_train_part_31',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_32\\\\dfdc_train_part_32',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_33\\\\dfdc_train_part_33',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_34\\\\dfdc_train_part_34',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_35\\\\dfdc_train_part_35',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_36\\\\dfdc_train_part_36',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_37\\\\dfdc_train_part_37',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_38\\\\dfdc_train_part_38',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_39\\\\dfdc_train_part_39',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_40\\\\dfdc_train_part_40',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_41\\\\dfdc_train_part_41',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_42\\\\dfdc_train_part_42',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_43\\\\dfdc_train_part_43',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_44\\\\dfdc_train_part_44',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_45\\\\dfdc_train_part_45',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_46\\\\dfdc_train_part_46',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_47\\\\dfdc_train_part_47',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_48\\\\dfdc_train_part_48',\n",
       " '../deepfake_train_full\\\\dfdc_train_part_49\\\\dfdc_train_part_49']"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_video_folders[9:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-29T11:06:34.283262Z",
     "start_time": "2019-12-29T11:06:34.278267Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4574"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(face_coordinates_per_video.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-28T11:11:28.763908Z",
     "start_time": "2019-12-28T11:11:28.759904Z"
    }
   },
   "outputs": [],
   "source": [
    "save_obj(face_coordinates_per_video, 'face_coordinates_per_video', '../face_detection/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-28T11:12:23.643555Z",
     "start_time": "2019-12-28T11:12:23.631552Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'aagfhgtpmv.mp4': {'0_0': {'coordinates': {'x1': 338,\n",
       "    'y1': 167,\n",
       "    'width': 219,\n",
       "    'height': 291},\n",
       "   'keypoints': {'left_eye': (416, 279),\n",
       "    'right_eye': (514, 266),\n",
       "    'nose': (495, 316),\n",
       "    'mouth_left': (452, 393),\n",
       "    'mouth_right': (527, 384)},\n",
       "   'confidence': 0.9999223947525024},\n",
       "  '75_0': {'coordinates': {'x1': 851, 'y1': 202, 'width': 198, 'height': 267},\n",
       "   'keypoints': {'left_eye': (900, 306),\n",
       "    'right_eye': (992, 303),\n",
       "    'nose': (943, 349),\n",
       "    'mouth_left': (910, 412),\n",
       "    'mouth_right': (985, 410)},\n",
       "   'confidence': 0.9999959468841553},\n",
       "  '150_0': {'coordinates': {'x1': 1020,\n",
       "    'y1': 189,\n",
       "    'width': 192,\n",
       "    'height': 275},\n",
       "   'keypoints': {'left_eye': (1058, 292),\n",
       "    'right_eye': (1146, 294),\n",
       "    'nose': (1081, 339),\n",
       "    'mouth_left': (1056, 396),\n",
       "    'mouth_right': (1132, 402)},\n",
       "   'confidence': 0.9999614953994751}}}"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "load_obj('face_coordinates_per_video', '../face_detection/')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## Other"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-27T18:05:55.708142Z",
     "start_time": "2019-12-27T18:05:02.738708Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aagfhgtpmv.mp4\n"
     ]
    },
    {
     "ename": "OSError",
     "evalue": "Could not load meta information\n=== stderr ===\n\n",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-117-af7292d1a233>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      2\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mfake_video_file_name\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mdf\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mdf\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'label'\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'FAKE'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'index'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mto_numpy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m     \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfake_video_file_name\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 4\u001b[1;33m     \u001b[0mvideo\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mVideo\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mos\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvideos_folder\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfake_video_file_name\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      5\u001b[0m     \u001b[0mfaces_from_video\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mFacesFromVideo\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvideo\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mvideo\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mframe_interval\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m25\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtotal_frames\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m5\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      6\u001b[0m     \u001b[1;32mbreak\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-87-0175b86cd53d>\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, path)\u001b[0m\n\u001b[0;32m      2\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__init__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpath\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpath\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpath\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 4\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontainer\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mimageio\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_reader\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'ffmpeg'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      5\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlength\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontainer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcount_frames\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      6\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfps\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcontainer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_meta_data\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'fps'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\core\\functions.py\u001b[0m in \u001b[0;36mget_reader\u001b[1;34m(uri, format, mode, **kwargs)\u001b[0m\n\u001b[0;32m    184\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    185\u001b[0m     \u001b[1;31m# Return its reader object\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 186\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0mformat\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_reader\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrequest\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    187\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    188\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\core\\format.py\u001b[0m in \u001b[0;36mget_reader\u001b[1;34m(self, request)\u001b[0m\n\u001b[0;32m    162\u001b[0m                 \u001b[1;34m\"Format %s cannot read in mode %r\"\u001b[0m \u001b[1;33m%\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mname\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mselect_mode\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    163\u001b[0m             )\n\u001b[1;32m--> 164\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mReader\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrequest\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    165\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    166\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mget_writer\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrequest\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\core\\format.py\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, format, request)\u001b[0m\n\u001b[0;32m    212\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_request\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mrequest\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    213\u001b[0m             \u001b[1;31m# Open the reader/writer\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 214\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_open\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m**\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrequest\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcopy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    215\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    216\u001b[0m         \u001b[1;33m@\u001b[0m\u001b[0mproperty\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\plugins\\ffmpeg.py\u001b[0m in \u001b[0;36m_open\u001b[1;34m(self, loop, size, dtype, pixelformat, print_info, ffmpeg_params, input_params, output_params, fps)\u001b[0m\n\u001b[0;32m    321\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    322\u001b[0m             \u001b[1;31m# Start ffmpeg subprocess and get meta information\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 323\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_initialize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    324\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    325\u001b[0m             \u001b[1;31m# For cameras, create thread that keeps reading the images\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio\\plugins\\ffmpeg.py\u001b[0m in \u001b[0;36m_initialize\u001b[1;34m(self, index)\u001b[0m\n\u001b[0;32m    464\u001b[0m                     \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_meta\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mupdate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmeta\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    465\u001b[0m             \u001b[1;32melif\u001b[0m \u001b[0mindex\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 466\u001b[1;33m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_meta\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mupdate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_read_gen\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__next__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    467\u001b[0m             \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    468\u001b[0m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_read_gen\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__next__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# we already have meta data\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\imageio_ffmpeg\\_io.py\u001b[0m in \u001b[0;36mread_frames\u001b[1;34m(path, pix_fmt, bpp, input_params, output_params)\u001b[0m\n\u001b[0;32m    148\u001b[0m             \u001b[0merr2\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlog_catcher\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_text\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0.2\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    149\u001b[0m             \u001b[0mfmt\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m\"Could not load meta information\\n=== stderr ===\\n{}\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 150\u001b[1;33m             \u001b[1;32mraise\u001b[0m \u001b[0mIOError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfmt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0merr2\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    151\u001b[0m         \u001b[1;32melif\u001b[0m \u001b[1;34m\"No such file or directory\"\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mlog_catcher\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mheader\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    152\u001b[0m             \u001b[1;32mraise\u001b[0m \u001b[0mIOError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"{} not found! Wrong path?\"\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mOSError\u001b[0m: Could not load meta information\n=== stderr ===\n\n"
     ]
    }
   ],
   "source": [
    "crops = {}\n",
    "for fake_video_file_name in df[df['label'] == 'FAKE']['index'].to_numpy():\n",
    "    print(fake_video_file_name)\n",
    "    video = Video(os.path.join(videos_folder, fake_video_file_name))\n",
    "    faces_from_video = FacesFromVideo(video=video, frame_interval=25, total_frames=5)\n",
    "    break\n",
    "    \n",
    "    \n",
    "    \n",
    "#     faces_from_video = extract_faces_from_video(video, 25, max_frames=20)\n",
    "#     save_crops('../to_be_annotated_fake/', fake_video_file_name.split('.')[0], faces_from_video)\n",
    "#     remove_img_from_faces_from_video(faces_from_video)\n",
    "#     crops[fake_video_file_name] = faces_from_video\n",
    "# save_obj(crops, 'crops')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-27T17:06:51.165868Z",
     "start_time": "2019-12-27T17:06:51.161866Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_path = '../deepfake_train_full'\n",
    "train_folders = [os.path.join(train_path, x) for x in os.listdir(train_path) if x[-1] != 'p']\n",
    "train_video_folders = [os.path.join(x, os.listdir(x)[0]) for x in train_folders]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-27T17:18:01.600420Z",
     "start_time": "2019-12-27T17:18:01.592423Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "train_folders = [os.path.join(train_path, x) for x in os.listdir(train_path) if x[-1] != 'p']\n",
    "train_video_folders = [os.path.join(x, os.listdir(x)[0]) for x in train_folders]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-27T17:18:05.065540Z",
     "start_time": "2019-12-27T17:18:05.061541Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'../deepfake_train_full\\\\dfdc_train_part_00\\\\dfdc_train_part_0'"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_video_folders[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-27T17:18:09.879540Z",
     "start_time": "2019-12-27T17:18:09.863538Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['aaqaifqrwn.mp4',\n",
       " 'aayrffkzxn.mp4',\n",
       " 'abhggqdift.mp4',\n",
       " 'acagallncj.mp4',\n",
       " 'acdkfksyev.mp4',\n",
       " 'achdeirhym.mp4',\n",
       " 'acryukkddn.mp4',\n",
       " 'adftmlvzfs.mp4',\n",
       " 'aegvtnwqoe.mp4',\n",
       " 'aejvkfbtxs.mp4',\n",
       " 'aeqpxjlbwu.mp4',\n",
       " 'ahhdprhoww.mp4',\n",
       " 'aimzesksew.mp4',\n",
       " 'aiuimrebtg.mp4',\n",
       " 'ajeegjgyzk.mp4',\n",
       " 'akpuczgfpk.mp4',\n",
       " 'alhqaurncj.mp4',\n",
       " 'alidfjtwvg.mp4',\n",
       " 'alnkzqihau.mp4',\n",
       " 'alpkghdgrv.mp4',\n",
       " 'ambabjrwbt.mp4',\n",
       " 'amluexcfvi.mp4',\n",
       " 'ansbtntiqu.mp4',\n",
       " 'aoydktojny.mp4',\n",
       " 'apedduehoy.mp4',\n",
       " 'apmzxocjio.mp4',\n",
       " 'apvzjkvnwn.mp4',\n",
       " 'apzckowxpy.mp4',\n",
       " 'arwyjgnvlp.mp4',\n",
       " 'asdzplbhnm.mp4',\n",
       " 'asxhhsants.mp4',\n",
       " 'atpngxfnhr.mp4',\n",
       " 'atrteuctxr.mp4',\n",
       " 'atzwrsivnb.mp4',\n",
       " 'augowetyej.mp4',\n",
       " 'augwyjvlqa.mp4',\n",
       " 'auquqlmsts.mp4',\n",
       " 'avoqheikrk.mp4',\n",
       " 'avznxuwvbx.mp4',\n",
       " 'awfjdqqygu.mp4',\n",
       " 'awigxxbbpb.mp4',\n",
       " 'awnfpubqmo.mp4',\n",
       " 'axbvoepbdc.mp4',\n",
       " 'axfhbpkdlc.mp4',\n",
       " 'axzjovaktp.mp4',\n",
       " 'aylqthgnhe.mp4',\n",
       " 'aymlynzeni.mp4',\n",
       " 'baiyzipjbr.mp4',\n",
       " 'bbnesewcot.mp4',\n",
       " 'bbxqtnztyy.mp4',\n",
       " 'bdnoxibdvz.mp4',\n",
       " 'bemkbovwsz.mp4',\n",
       " 'bfktkdugru.mp4',\n",
       " 'bhvnfaylpk.mp4',\n",
       " 'biatmogsbv.mp4',\n",
       " 'bihlajjbnl.mp4',\n",
       " 'birmbtezni.mp4',\n",
       " 'bjxasekfcx.mp4',\n",
       " 'bkcztuqxbp.mp4',\n",
       " 'bldeizqycp.mp4',\n",
       " 'blwvnvsbmz.mp4',\n",
       " 'bmfkkzheii.mp4',\n",
       " 'bmhghysucl.mp4',\n",
       " 'bnitzihcqs.mp4',\n",
       " 'bobihhgccf.mp4',\n",
       " 'bpguesjrfa.mp4',\n",
       " 'bqamrsyvnf.mp4',\n",
       " 'bqefygjsnx.mp4',\n",
       " 'bqvhlqlrba.mp4',\n",
       " 'brihlutndc.mp4',\n",
       " 'brpkkgdggv.mp4',\n",
       " 'brtujopkby.mp4',\n",
       " 'btizoxwqwp.mp4',\n",
       " 'btyxvpjgcc.mp4',\n",
       " 'buzdtxdyid.mp4',\n",
       " 'bwjqcwhpfr.mp4',\n",
       " 'bwlxucgxro.mp4',\n",
       " 'bwsytelwkm.mp4',\n",
       " 'bxdhnnffvw.mp4',\n",
       " 'bxgbpdjgjv.mp4',\n",
       " 'bxooifrfzo.mp4',\n",
       " 'bykgnyykta.mp4',\n",
       " 'bytfrsxnwi.mp4',\n",
       " 'bztdemptfg.mp4',\n",
       " 'bzvzpwrabw.mp4',\n",
       " 'bzxauvpscl.mp4',\n",
       " 'bzyawroeif.mp4',\n",
       " 'calhtslbzf.mp4',\n",
       " 'cbybbxuusl.mp4',\n",
       " 'cbzxykepbl.mp4',\n",
       " 'ccjiwxnihq.mp4',\n",
       " 'ccrrlidqnp.mp4',\n",
       " 'cdcdoqbdmi.mp4',\n",
       " 'cdyotmvujc.mp4',\n",
       " 'cefpmcqwex.mp4',\n",
       " 'cemftmsxis.mp4',\n",
       " 'cerwqpmouj.mp4',\n",
       " 'cfhejdmemu.mp4',\n",
       " 'cfyayetydv.mp4',\n",
       " 'chqqxfuuzi.mp4',\n",
       " 'chzqptmqxv.mp4',\n",
       " 'citnmpozjl.mp4',\n",
       " 'cjhwsuzeuz.mp4',\n",
       " 'cjndhcclqi.mp4',\n",
       " 'cjxbcxpbqi.mp4',\n",
       " 'ckfqqhpbph.mp4',\n",
       " 'clpvgmopjx.mp4',\n",
       " 'clzkcmoyhb.mp4',\n",
       " 'cmoizyxqcc.mp4',\n",
       " 'cmvymfbmlb.mp4',\n",
       " 'cogcvbbjcm.mp4',\n",
       " 'coujjnypba.mp4',\n",
       " 'cqbhgsmffd.mp4',\n",
       " 'cqhwesrciw.mp4',\n",
       " 'cqomufbzpp.mp4',\n",
       " 'cqpdqjzgcs.mp4',\n",
       " 'cqroxivqio.mp4',\n",
       " 'cqtxrekabh.mp4',\n",
       " 'crflaflooj.mp4',\n",
       " 'crwzrpozbv.mp4',\n",
       " 'csofafwrts.mp4',\n",
       " 'csqojsqtgn.mp4',\n",
       " 'cthdrbnhpo.mp4',\n",
       " 'ctooqkhtgp.mp4',\n",
       " 'ctzgudcdyu.mp4',\n",
       " 'culaevxduf.mp4',\n",
       " 'cvboojbsxa.mp4',\n",
       " 'cvdrzlkwdf.mp4',\n",
       " 'cvmcvcqpgn.mp4',\n",
       " 'cvqyddpcxj.mp4',\n",
       " 'cvycgtwfsi.mp4',\n",
       " 'cwqycdjevy.mp4',\n",
       " 'cwvpcctjgs.mp4',\n",
       " 'cxgieqcozm.mp4',\n",
       " 'cxgsjamupo.mp4',\n",
       " 'cxjtjcinqt.mp4',\n",
       " 'cybnhobdyh.mp4',\n",
       " 'cygixvotoj.mp4',\n",
       " 'cyhxxnezbo.mp4',\n",
       " 'cyifdylbln.mp4',\n",
       " 'cythdgxpdi.mp4',\n",
       " 'dabysiasvf.mp4',\n",
       " 'dahvuijxob.mp4',\n",
       " 'dbgztzhbhb.mp4',\n",
       " 'dbhthsgych.mp4',\n",
       " 'dcwnkxrsgg.mp4',\n",
       " 'ddqtkrkmjn.mp4',\n",
       " 'ddtbarpcgo.mp4',\n",
       " 'demuhxssgl.mp4',\n",
       " 'dfcoyuhhex.mp4',\n",
       " 'dflpzzixje.mp4',\n",
       " 'dghhdvdzrb.mp4',\n",
       " 'dgyeaodnzw.mp4',\n",
       " 'dhjnjkzuhq.mp4',\n",
       " 'dhyybsrvwc.mp4',\n",
       " 'dibwmyrpql.mp4',\n",
       " 'didzujjhtg.mp4',\n",
       " 'dilfhyevdv.mp4',\n",
       " 'djgbahezbh.mp4',\n",
       " 'dkenctaoes.mp4',\n",
       " 'dkojxwpvza.mp4',\n",
       " 'dlfuxshert.mp4',\n",
       " 'dlpaibcwsy.mp4',\n",
       " 'dlrcprykef.mp4',\n",
       " 'dlybulufda.mp4',\n",
       " 'dmirtsrziw.mp4',\n",
       " 'doiljuhamq.mp4',\n",
       " 'dolynaukae.mp4',\n",
       " 'doniqevxeg.mp4',\n",
       " 'dootttqtvv.mp4',\n",
       " 'dpevefkefv.mp4',\n",
       " 'dqlmffnywx.mp4',\n",
       " 'drjwafunon.mp4',\n",
       " 'drlkcilgmc.mp4',\n",
       " 'drmtvuyoam.mp4',\n",
       " 'drooaxkvib.mp4',\n",
       " 'dsgswagtvp.mp4',\n",
       " 'dsiirbwwyt.mp4',\n",
       " 'dsvopuhqiz.mp4',\n",
       " 'dtfyluocuh.mp4',\n",
       " 'dtjcyzgdts.mp4',\n",
       " 'dtmlspmqqp.mp4',\n",
       " 'dtwxxkuvox.mp4',\n",
       " 'dtwzjhoizi.mp4',\n",
       " 'duciovvndy.mp4',\n",
       " 'duczhwighk.mp4',\n",
       " 'dvrnrzlqdo.mp4',\n",
       " 'dwkkqhxclz.mp4',\n",
       " 'dxdzrfikon.mp4',\n",
       " 'dxfdovivlw.mp4',\n",
       " 'dxrqdhegop.mp4',\n",
       " 'dyjklprkoc.mp4',\n",
       " 'dypesfpwyi.mp4',\n",
       " 'dywronddes.mp4',\n",
       " 'dzlpfpszyv.mp4',\n",
       " 'dzpokpqpxr.mp4',\n",
       " 'ealulizawd.mp4',\n",
       " 'eauhakbohz.mp4',\n",
       " 'ebzsqgfpgh.mp4',\n",
       " 'echdilwfig.mp4',\n",
       " 'eckloytegm.mp4',\n",
       " 'ecnmwevrxo.mp4',\n",
       " 'edqhdckynd.mp4',\n",
       " 'eefzpdrrji.mp4',\n",
       " 'efktbhpilh.mp4',\n",
       " 'ehfmarmsvo.mp4',\n",
       " 'eilfbdisof.mp4',\n",
       " 'eisofhptvk.mp4',\n",
       " 'eiwtggvtfp.mp4',\n",
       " 'ejhhdlzpjg.mp4',\n",
       " 'ekkheaisuo.mp4',\n",
       " 'elxflqcpgp.mp4',\n",
       " 'embmaestpq.mp4',\n",
       " 'eneofiltbk.mp4',\n",
       " 'enhiqdvfpu.mp4',\n",
       " 'enqxsifggs.mp4',\n",
       " 'eowsoapzuj.mp4',\n",
       " 'epeyujjlhn.mp4',\n",
       " 'eppmwpntfi.mp4',\n",
       " 'eppyqpgewp.mp4',\n",
       " 'epwesrdmpt.mp4',\n",
       " 'eqakbrnwmd.mp4',\n",
       " 'erbfxwbqic.mp4',\n",
       " 'erbgcpjuzj.mp4',\n",
       " 'ernfzaruwz.mp4',\n",
       " 'escwylyxap.mp4',\n",
       " 'esjdyghhog.mp4',\n",
       " 'espkiocpxq.mp4',\n",
       " 'etdliwticv.mp4',\n",
       " 'etychryvty.mp4',\n",
       " 'euavjxseoz.mp4',\n",
       " 'eugcprxhvd.mp4',\n",
       " 'eurhttghny.mp4',\n",
       " 'euzlhrwolz.mp4',\n",
       " 'ewknjjmzqb.mp4',\n",
       " 'exseruhiuk.mp4',\n",
       " 'exxqlfpnbz.mp4',\n",
       " 'eyyigsrqff.mp4',\n",
       " 'ezzfdnllnp.mp4',\n",
       " 'fahrejqccb.mp4',\n",
       " 'faktmhmtpb.mp4',\n",
       " 'fbgfjwjrzl.mp4',\n",
       " 'fbxpssbrqp.mp4',\n",
       " 'fckxaqjbxk.mp4',\n",
       " 'fdpisghkmd.mp4',\n",
       " 'fdqafjiqtn.mp4',\n",
       " 'fdzrcjpewd.mp4',\n",
       " 'fecaapkwdm.mp4',\n",
       " 'fedcdqtboz.mp4',\n",
       " 'fgbopxbkam.mp4',\n",
       " 'fgobmbcami.mp4',\n",
       " 'fhgnmqzfca.mp4',\n",
       " 'fhtmnfeqft.mp4',\n",
       " 'fhzdtfiqrh.mp4',\n",
       " 'fjafyicjlw.mp4',\n",
       " 'fjrueenjyp.mp4',\n",
       " 'flilazlokf.mp4',\n",
       " 'fmeptfderu.mp4',\n",
       " 'fnkkqtouoz.mp4',\n",
       " 'fnslimfagb.mp4',\n",
       " 'fokjbuazgr.mp4',\n",
       " 'fonrexmbzz.mp4',\n",
       " 'fopjiyxiqd.mp4',\n",
       " 'fpcncmwzxf.mp4',\n",
       " 'fqmfvshgns.mp4',\n",
       " 'frmzkdhkzw.mp4',\n",
       " 'frqhofmoob.mp4',\n",
       " 'frsjawvaoq.mp4',\n",
       " 'fsaronfupy.mp4',\n",
       " 'fsdrwikhge.mp4',\n",
       " 'fsiunpwpqx.mp4',\n",
       " 'fswszwicfw.mp4',\n",
       " 'fufcmupzen.mp4',\n",
       " 'fuiglusbnl.mp4',\n",
       " 'furvgnwfln.mp4',\n",
       " 'fuzdxjsqvp.mp4',\n",
       " 'fxredodlkl.mp4',\n",
       " 'fxuxxtryjn.mp4',\n",
       " 'fybkxkgstv.mp4',\n",
       " 'fydhhlvjsj.mp4',\n",
       " 'fyrckgsqvg.mp4',\n",
       " 'fzkzqwoghs.mp4',\n",
       " 'fzqnoyzhbm.mp4',\n",
       " 'fzvpbrzssi.mp4',\n",
       " 'gbatfzbsjs.mp4',\n",
       " 'gbgtzspdnv.mp4',\n",
       " 'gccnvdoknm.mp4',\n",
       " 'gcdtglsoqj.mp4',\n",
       " 'gcnhqdxwqd.mp4',\n",
       " 'gcphojsohr.mp4',\n",
       " 'gczieafvpe.mp4',\n",
       " 'gecrjewspu.mp4',\n",
       " 'geknowufwz.mp4',\n",
       " 'gepzcywmsx.mp4',\n",
       " 'getmctvhmd.mp4',\n",
       " 'gfcycflhbo.mp4',\n",
       " 'gfdjzwnpyp.mp4',\n",
       " 'gfkdvulbfo.mp4',\n",
       " 'gfknvvudwk.mp4',\n",
       " 'gfncmhgdio.mp4',\n",
       " 'gfoozygynw.mp4',\n",
       " 'gilzdnbpep.mp4',\n",
       " 'gjiziopvis.mp4',\n",
       " 'gjjpzsgigh.mp4',\n",
       " 'gjxfowkaeo.mp4',\n",
       " 'gkicllwmyb.mp4',\n",
       " 'gkqitjodaa.mp4',\n",
       " 'gldlvkplzy.mp4',\n",
       " 'gmaghrdamw.mp4',\n",
       " 'gmctmgmyob.mp4',\n",
       " 'gmonvwaujc.mp4',\n",
       " 'gnlvnyrpfq.mp4',\n",
       " 'gnmmhlbzge.mp4',\n",
       " 'gnslnjrgnx.mp4',\n",
       " 'gnxgiwcknq.mp4',\n",
       " 'gnxlbuwucy.mp4',\n",
       " 'gohsbddbhu.mp4',\n",
       " 'goxbkycgkg.mp4',\n",
       " 'goyrhrvden.mp4',\n",
       " 'gozbrlxguv.mp4',\n",
       " 'gpjhljtyoq.mp4',\n",
       " 'gpskwigkxd.mp4',\n",
       " 'gpzwvdkuie.mp4',\n",
       " 'gqnaxievjx.mp4',\n",
       " 'gqnkarjtmh.mp4',\n",
       " 'grqyyhfexr.mp4',\n",
       " 'gtdjfweoyr.mp4',\n",
       " 'gtfwzfrxml.mp4',\n",
       " 'gtinxlhder.mp4',\n",
       " 'gudedkpnie.mp4',\n",
       " 'gujnhuvarb.mp4',\n",
       " 'gumgvhycum.mp4',\n",
       " 'guqayoypew.mp4',\n",
       " 'gvasarkpfh.mp4',\n",
       " 'gvbcwuyaot.mp4',\n",
       " 'gvcrwbnzpa.mp4',\n",
       " 'gvlraqlweh.mp4',\n",
       " 'gwsmeakcic.mp4',\n",
       " 'gwzfugbidx.mp4',\n",
       " 'gxonqiyzpk.mp4',\n",
       " 'gxzdueambi.mp4',\n",
       " 'gyosvoeamf.mp4',\n",
       " 'gyrdgpiygb.mp4',\n",
       " 'gzkggvloao.mp4',\n",
       " 'hasjjkoygr.mp4',\n",
       " 'haxnhfvqvs.mp4',\n",
       " 'hbdvyzpqny.mp4',\n",
       " 'hbtcuqhwyy.mp4',\n",
       " 'hcanfkwivl.mp4',\n",
       " 'hclsparpth.mp4',\n",
       " 'hcqxrwzgxt.mp4',\n",
       " 'hdgfgvcoqk.mp4',\n",
       " 'hdytjaiglk.mp4',\n",
       " 'hectzcuzdv.mp4',\n",
       " 'hefisnapds.mp4',\n",
       " 'heiyoojifp.mp4',\n",
       " 'hfeacozznh.mp4',\n",
       " 'hgjdosshpt.mp4',\n",
       " 'hhddxuhorq.mp4',\n",
       " 'hhyxweyjlt.mp4',\n",
       " 'hicjuubiau.mp4',\n",
       " 'hitfycdavv.mp4',\n",
       " 'hivnldfvyl.mp4',\n",
       " 'hjddjiqcpb.mp4',\n",
       " 'hjxckindzx.mp4',\n",
       " 'hkvcbcfulf.mp4',\n",
       " 'hldhchyhfg.mp4',\n",
       " 'hnalkvbcim.mp4',\n",
       " 'horutnradk.mp4',\n",
       " 'hphiyksrth.mp4',\n",
       " 'hplxtssgnz.mp4',\n",
       " 'hpwdlgshqg.mp4',\n",
       " 'hqcrauawma.mp4',\n",
       " 'hqxfenmiis.mp4',\n",
       " 'hqzwudvhih.mp4',\n",
       " 'hrbkraauwx.mp4',\n",
       " 'hrovzkcyjw.mp4',\n",
       " 'hsbljbsgxr.mp4',\n",
       " 'hsbrwaeaah.mp4',\n",
       " 'hsbwhlolsn.mp4',\n",
       " 'hsosrxlnzk.mp4',\n",
       " 'hsvbimzuza.mp4',\n",
       " 'hsxrmrcaqh.mp4',\n",
       " 'hthuorhwdn.mp4',\n",
       " 'htorvhbcae.mp4',\n",
       " 'hurcutcxpb.mp4',\n",
       " 'huznbaakma.mp4',\n",
       " 'hvlqewtbbf.mp4',\n",
       " 'hvqfhieetb.mp4',\n",
       " 'hwcnsgqzld.mp4',\n",
       " 'hwibtynenh.mp4',\n",
       " 'hwtupajxgl.mp4',\n",
       " 'hyjqolupxn.mp4',\n",
       " 'hzmppzxsuy.mp4',\n",
       " 'iafvzgpbix.mp4',\n",
       " 'iagkfqsiqf.mp4',\n",
       " 'iazovjywij.mp4',\n",
       " 'ibjomudxdp.mp4',\n",
       " 'idczhqbbqz.mp4',\n",
       " 'idmruoiylw.mp4',\n",
       " 'idsupeszcp.mp4',\n",
       " 'ielrrtwsdr.mp4',\n",
       " 'iffvaaswwj.mp4',\n",
       " 'ifjzcxxvcd.mp4',\n",
       " 'iftluclqfd.mp4',\n",
       " 'ihnhxathkq.mp4',\n",
       " 'ihrjvevuce.mp4',\n",
       " 'iimnvjpgoy.mp4',\n",
       " 'ijehzkozkp.mp4',\n",
       " 'ijgcwhjkua.mp4',\n",
       " 'ikjlhrwzeo.mp4',\n",
       " 'ikpyxzpiiv.mp4',\n",
       " 'ilflslfgbn.mp4',\n",
       " 'imndwwlfui.mp4',\n",
       " 'inwiunnxmr.mp4',\n",
       " 'inxftbkcqq.mp4',\n",
       " 'inzrlbgtul.mp4',\n",
       " 'iocjgjemeb.mp4',\n",
       " 'iodwjkqblp.mp4',\n",
       " 'iojmxrzvxe.mp4',\n",
       " 'ioolqahamw.mp4',\n",
       " 'iorrnxiyyo.mp4',\n",
       " 'irlgidcrvj.mp4',\n",
       " 'isgjlgucub.mp4',\n",
       " 'itfsvvmslp.mp4',\n",
       " 'itwvssoevj.mp4',\n",
       " 'iugnysiwao.mp4',\n",
       " 'iurmztyqed.mp4',\n",
       " 'ivdlxktqvk.mp4',\n",
       " 'ivtbbwejej.mp4',\n",
       " 'iwswulwtjp.mp4',\n",
       " 'iwtjnrbcbb.mp4',\n",
       " 'iyhkhsdhma.mp4',\n",
       " 'iyhzzwoqmi.mp4',\n",
       " 'iyswglziib.mp4',\n",
       " 'izjvqhjokl.mp4',\n",
       " 'jakregbmuf.mp4',\n",
       " 'jaurxwocly.mp4',\n",
       " 'jawgcggquk.mp4',\n",
       " 'jbwtvoixgc.mp4',\n",
       " 'jbxjvkdnci.mp4',\n",
       " 'jbxviammiv.mp4',\n",
       " 'jccvzepxjn.mp4',\n",
       " 'jdfaxqtpjk.mp4',\n",
       " 'jdkvgcjxje.mp4',\n",
       " 'jdzgjbqksy.mp4',\n",
       " 'jeblbsmlij.mp4',\n",
       " 'jeosukfzow.mp4',\n",
       " 'jfvsrcfame.mp4',\n",
       " 'jgdzqmqalr.mp4',\n",
       " 'jgybxxtltz.mp4',\n",
       " 'jgzwhkcrcc.mp4',\n",
       " 'jhczqfefgw.mp4',\n",
       " 'jhetgmikln.mp4',\n",
       " 'jinomojicv.mp4',\n",
       " 'jjguwagxur.mp4',\n",
       " 'jjxzidyitl.mp4',\n",
       " 'jkfoehpdnd.mp4',\n",
       " 'jkqivqkrnc.mp4',\n",
       " 'jkxjvuioek.mp4',\n",
       " 'jmcryhnksj.mp4',\n",
       " 'jmhqbsgzrg.mp4',\n",
       " 'jmkjrabiak.mp4',\n",
       " 'jmkqtjuuqu.mp4',\n",
       " 'jorfifqxhq.mp4',\n",
       " 'jphmkhpoeb.mp4',\n",
       " 'jppzzcldui.mp4',\n",
       " 'jqlsethxvz.mp4',\n",
       " 'jquevmhdvc.mp4',\n",
       " 'jqypqwufdv.mp4',\n",
       " 'jranrlmuyk.mp4',\n",
       " 'jrgfabamaa.mp4',\n",
       " 'jsbpkpxwew.mp4',\n",
       " 'jstwvimaml.mp4',\n",
       " 'jtpupxzgoa.mp4',\n",
       " 'jtrbieopvk.mp4',\n",
       " 'juzjntqofi.mp4',\n",
       " 'jvadnjprqn.mp4',\n",
       " 'jvtjxreizj.mp4',\n",
       " 'jwkgjrmfxp.mp4',\n",
       " 'jxlhrgwjdt.mp4',\n",
       " 'jxqibxwhlc.mp4',\n",
       " 'jxvoarslil.mp4',\n",
       " 'jytrvwlewz.mp4',\n",
       " 'jzupayeuln.mp4',\n",
       " 'kadzgvxsii.mp4',\n",
       " 'kafmsctues.mp4',\n",
       " 'kajzbtxqdk.mp4',\n",
       " 'kanuhxwqaa.mp4',\n",
       " 'kbssbcdcxx.mp4',\n",
       " 'kcidxranvr.mp4',\n",
       " 'kdiokpezzi.mp4',\n",
       " 'kdycxkmnnw.mp4',\n",
       " 'kdzirwrnyk.mp4',\n",
       " 'kedahvzait.mp4',\n",
       " 'keuvybqsbr.mp4',\n",
       " 'kezvvweubg.mp4',\n",
       " 'kfbdnasciu.mp4',\n",
       " 'kfgdvqjuzu.mp4',\n",
       " 'kfhmhoxjtx.mp4',\n",
       " 'kfnrvxndoj.mp4',\n",
       " 'kgcbejdwal.mp4',\n",
       " 'kgfcupxgur.mp4',\n",
       " 'khfyqlxazq.mp4',\n",
       " 'khtwrijuqn.mp4',\n",
       " 'kihrpdtbwy.mp4',\n",
       " 'kiswvzlayw.mp4',\n",
       " 'kiyuduxzuc.mp4',\n",
       " 'kjpxvgjsal.mp4',\n",
       " 'kjwnylvwuo.mp4',\n",
       " 'kjzyrkhnaf.mp4',\n",
       " 'kkemhrscxz.mp4',\n",
       " 'kkhrtmftlq.mp4',\n",
       " 'kkxgqvpusd.mp4',\n",
       " 'kmcdjxmnoa.mp4',\n",
       " 'kmhiqddyhh.mp4',\n",
       " 'knxltsvzyu.mp4',\n",
       " 'komngcqveq.mp4',\n",
       " 'koxwykmawn.mp4',\n",
       " 'kpdsblcqvg.mp4',\n",
       " 'kpmieleslf.mp4',\n",
       " 'kpwwhpqfkx.mp4',\n",
       " 'kqlvggiqee.mp4',\n",
       " 'kqsjsvinkd.mp4',\n",
       " 'kquembvlju.mp4',\n",
       " 'ksjpgrujqv.mp4',\n",
       " 'kssj"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<b>limit_output extension: Maximum message size of 10000 exceeded with 19004 characters</b>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "os.listdir(train_video_folders[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## Get faces from videos using CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T14:01:28.391717Z",
     "start_time": "2019-12-26T14:01:28.373741Z"
    },
    "code_folding": [
     17
    ],
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# extract a single face from a given photograph\n",
    "def extract_faces_from_video(video, frame_interval, max_frames=None):\n",
    "    faces_from_video = {}\n",
    "    stop = video.__len__() if max_frames == None else max_frames*frame_interval\n",
    "    stop = stop if stop <= video.__len__() else video.__len__()\n",
    "    for frame_nr in range(0, stop, frame_interval):\n",
    "        img = video.get(frame_nr)\n",
    "        faces = extract_faces_from_img(img, frame_nr)\n",
    "        faces_from_video.update(faces)\n",
    "        \n",
    "    # faces_from_video = filter_true_faces(faces_from_video)\n",
    "    return faces_from_video\n",
    "\n",
    "def get_max_faces_per_video(faces_from_video):\n",
    "    return len(np.unique(np.array([int(x.split('_')[1]) for x in faces_from_video.keys()])))\n",
    "\n",
    "def filter_true_faces(faces_from_video):\n",
    "    max_faces_per_video = get_max_faces_per_video(faces_from_video)\n",
    "    if max_faces_per_video == 1:\n",
    "        return max_faces_per_video\n",
    "    # Coordinates into dict\n",
    "    coordinates = {}\n",
    "    for key in  faces_from_video:\n",
    "        coordinates[key] = []\n",
    "        coordinates[key].append(faces_from_video[key]['coordinates'])\n",
    "            \n",
    "    #  Coordinates into array\n",
    "    data = []\n",
    "    for key in coordinates.keys():\n",
    "        for c in coordinates[key]:\n",
    "            data.append([c['x1'], c['y1'], c['width'], c['height']])\n",
    "    x = np.array(data)\n",
    "    \n",
    "    # Cluster important images only\n",
    "    kmeans = KMeans(n_clusters=max_faces_per_video, random_state=0).fit(x)\n",
    "    u, c = np.unique(kmeans.labels_, return_counts=True)\n",
    "    y = u[c == c.max()]\n",
    "    if len(y) > 1:\n",
    "        print('multiple faces equally present')\n",
    "        return faces_from_video\n",
    "    else:\n",
    "        idx = np.where(kmeans.labels_ == y[0])[0]\n",
    "        faces_keys = np.array(list(faces_from_video.keys()))[list(idx)]\n",
    "        faces_from_video = { key: faces_from_video[key] for key in faces_keys }\n",
    "\n",
    "    return faces_from_video\n",
    "\n",
    "def extract_faces_from_img(img, frame_nr):\n",
    "    faces = {}\n",
    "    detector = MTCNN()\n",
    "    results = detector.detect_faces(img)\n",
    "    for idx, res in enumerate(results):\n",
    "        box = np.array(res['box'])\n",
    "        box[box < 0] = 0\n",
    "        x1, y1, width, height = box\n",
    "        x2, y2 = x1 + width, y1 + height\n",
    "        face = img[y1:y2, x1:x2]\n",
    "        image = Image.fromarray(face)\n",
    "        key = str(str(frame_nr) + '_' + str(idx))\n",
    "        if res['confidence'] > 0.95:\n",
    "            faces[key] = {}\n",
    "            faces[key]['img'] = image\n",
    "            faces[key]['coordinates'] = {'x1': x1, \n",
    "                                         'y1': y1, \n",
    "                                         'width': width, \n",
    "                                         'height': height}\n",
    "            faces[key]['keypoints'] = res['keypoints']\n",
    "            faces[key]['confidence'] = res['confidence']\n",
    "        \n",
    "    return faces\n",
    "\n",
    "def save_crops(path, video_file_name, crops):\n",
    "    os.makedirs(path, exist_ok=True)\n",
    "    for key in crops.keys():\n",
    "        crop = crops[key]['img']\n",
    "        imwrite(os.path.join(path, '{}_{}.{}'.format(video_file_name, key, 'png')), crop)\n",
    "\n",
    "def remove_img_from_faces_from_video(faces_from_video):\n",
    "    for key in faces_from_video.keys():\n",
    "        del faces_from_video[key]['img']\n",
    "        \n",
    "def save_obj(obj, name, fake=True):\n",
    "    fake_or_true = 'fake' if fake else 'real'\n",
    "    with open('../to_be_annotated_{}/'.format(fake_or_true) + name + '.pkl', 'wb') as f:\n",
    "        pickle.dump(obj, f, pickle.HIGHEST_PROTOCOL)\n",
    "\n",
    "def load_obj(name, fake=True):\n",
    "    fake_or_true = 'fake' if fake else 'real'\n",
    "    with open('../to_be_annotated_{}/'.format(fake_or_true) + name + '.pkl', 'rb') as f:\n",
    "        return pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T04:17:41.018437Z",
     "start_time": "2019-12-26T02:00:54.036126Z"
    },
    "deletable": false,
    "editable": false,
    "hidden": true,
    "run_control": {
     "frozen": true
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aagfhgtpmv.mp4\n",
      "aapnvogymq.mp4\n",
      "abofeumbvv.mp4\n",
      "abqwwspghj.mp4\n",
      "acifjvzvpm.mp4\n",
      "acqfdwsrhi.mp4\n",
      "acxnxvbsxk.mp4\n",
      "acxwigylke.mp4\n",
      "aczrgyricp.mp4\n",
      "adhsbajydo.mp4\n",
      "adohikbdaz.mp4\n",
      "adylbeequz.mp4\n",
      "aelzhcnwgf.mp4\n",
      "aettqgevhz.mp4\n",
      "aevrfsexku.mp4\n",
      "agdkmztvby.mp4\n",
      "agqphdxmwt.mp4\n",
      "ahbweevwpv.mp4\n",
      "ahdbuwqxit.mp4\n",
      "ahfazfbntc.mp4\n",
      "aipfdnwpoo.mp4\n",
      "ajwpjhrbcv.mp4\n",
      "aklqzsddfl.mp4\n",
      "aknbdpmgua.mp4\n",
      "aknmpoonls.mp4\n",
      "akvmwkdyuv.mp4\n",
      "akxoopqjqz.mp4\n",
      "akzbnazxtz.mp4\n",
      "aladcziidp.mp4\n",
      "alaijyygdv.mp4\n",
      "alninxcyhg.mp4\n",
      "altziddtxi.mp4\n",
      "alvgwypubw.mp4\n",
      "amaivqofda.mp4\n",
      "amowujxmzc.mp4\n",
      "andaxzscny.mp4\n",
      "aneclqfpbt.mp4\n",
      "aorjvbyxhw.mp4\n",
      "apatcsqejh.mp4\n",
      "apgjqzkoma.mp4\n",
      "apogckdfrz.mp4\n",
      "aqpnvjhuzw.mp4\n",
      "arkroixhey.mp4\n",
      "arlmiizoob.mp4\n",
      "arrhsnjqku.mp4\n",
      "asdpeebotb.mp4\n",
      "aslsvlvpth.mp4\n",
      "asmpfjfzif.mp4\n",
      "asvcrfdpnq.mp4\n",
      "atxvxouljq.mp4\n",
      "atyntldecu.mp4\n",
      "atzdznmder.mp4\n",
      "aufmsmnoye.mp4\n",
      "augtsuxpzc.mp4\n",
      "avfitoutyn.mp4\n",
      "avgiuextiz.mp4\n",
      "avibnnhwhp.mp4\n",
      "avnqydkqjj.mp4\n",
      "avssvvsdhz.mp4\n",
      "avtycwsgyb.mp4\n",
      "avvdgsennp.mp4\n",
      "avywawptfc.mp4\n",
      "awhmfnnjih.mp4\n",
      "awnwkrqibf.mp4\n",
      "awukslzjra.mp4\n",
      "axczxisdtb.mp4\n",
      "axoygtekut.mp4\n",
      "axwgcsyphv.mp4\n",
      "axwovszumc.mp4\n",
      "ayqvfdhslr.mp4\n",
      "azpuxunqyo.mp4\n",
      "azsmewqghg.mp4\n",
      "bahdpoesir.mp4\n",
      "bbhpvrmbse.mp4\n",
      "bbhtdfuqxq.mp4\n",
      "bbvgxeczei.mp4\n",
      "bchnbulevv.mp4\n",
      "bctvsmddgq.mp4\n",
      "bdbhekrrwo.mp4\n",
      "bdgipnyobr.mp4\n",
      "bdxuhamuqx.mp4\n",
      "benmsfzfaz.mp4\n",
      "bgaogsjehq.mp4\n",
      "bggsurpgpr.mp4\n",
      "bghphrsfxf.mp4\n",
      "bgmlwsoamc.mp4\n",
      "bguwlyazau.mp4\n",
      "bhaaboftbc.mp4\n",
      "bhbdugnurr.mp4\n",
      "bhpwpydzpo.mp4\n",
      "bhsluedavd.mp4\n",
      "bjjbwsqjir.mp4\n",
      "bjkmjilrxp.mp4\n",
      "bjsmaqefoi.mp4\n",
      "bkmdzhfzfh.mp4\n",
      "bkvetcojbt.mp4\n",
      "bkwxhglwct.mp4\n",
      "blpchvmhxx.mp4\n",
      "blzydqdfem.mp4\n",
      "bmbbkwmxqj.mp4\n",
      "bmehkyanbj.mp4\n",
      "bmhvktyiwp.mp4\n",
      "bmioepcpsx.mp4\n",
      "bmjmjmbglm.mp4\n",
      "bnbuonyoje.mp4\n",
      "bndybcqhfr.mp4\n",
      "bnjcdrfuov.mp4\n",
      "bntlodcfeg.mp4\n",
      "bofqajtwve.mp4\n",
      "boovltmuwi.mp4\n",
      "bopqhhalml.mp4\n",
      "bourlmzsio.mp4\n",
      "bpwzipqtxf.mp4\n",
      "bpxckdzddv.mp4\n",
      "bqdjzqhcft.mp4\n",
      "bqeiblbxtl.mp4\n",
      "bqhtpqmmqp.mp4\n",
      "bqkdbcqjvb.mp4\n",
      "bqnymlsayl.mp4\n",
      "bqqpbzjgup.mp4\n",
      "bqtuuwzdtr.mp4\n",
      "brhalypwoo.mp4\n",
      "brvqtabyxj.mp4\n",
      "bseamdrpbj.mp4\n",
      "bsfmwclnqy.mp4\n",
      "bsqgziaylx.mp4\n",
      "btiysiskpf.mp4\n",
      "btjlfpzbdu.mp4\n",
      "btjwbtsgln.mp4\n",
      "btmsngnqhv.mp4\n",
      "btohlidmru.mp4\n",
      "btugrnoton.mp4\n",
      "btunxncpjh.mp4\n",
      "btxlttbpkj.mp4\n",
      "bvgwelbeof.mp4\n",
      "bvzjkezkms.mp4\n",
      "bweezhfpzp.mp4\n",
      "bwuwstvsbw.mp4\n",
      "bydaidkpdp.mp4\n",
      "byfenovjnf.mp4\n",
      "byijojkdba.mp4\n",
      "byofowlkki.mp4\n",
      "byqzyxifza.mp4\n",
      "byunigvnay.mp4\n",
      "byyqectxqa.mp4\n",
      "bzmdrafeex.mp4\n",
      "caqbrkogkb.mp4\n",
      "cbbibzcoih.mp4\n",
      "cbltdtxglo.mp4\n",
      "ccmonzqfrz.mp4\n",
      "cdaxixbosp.mp4\n",
      "cdbsbdymzd.mp4\n",
      "cdphtzqrvp.mp4\n",
      "cdyakrxkia.mp4\n",
      "cepxysienc.mp4\n",
      "cettndmvzl.mp4\n",
      "ceymbecxnj.mp4\n",
      "cferslmfwh.mp4\n",
      "cffffbcywc.mp4\n",
      "cfyduhpbps.mp4\n",
      "cglxirfaey.mp4\n",
      "cgvrgibpfo.mp4\n",
      "chzieimrwu.mp4\n",
      "ckbdwedgmc.mp4\n",
      "cknyxaqouy.mp4\n",
      "cksanfsjhc.mp4\n",
      "clihsshdkq.mp4\n",
      "cmxcfkrjiv.mp4\n",
      "cnilkgvfei.mp4\n",
      "coadfnerlk.mp4\n",
      "covdcysmbi.mp4\n",
      "cqfugiqupm.mp4\n",
      "cqhngvpgyi.mp4\n",
      "cqrskwiqng.mp4\n",
      "crktehraph.mp4\n",
      "crzfebnfgb.mp4\n",
      "cthdnahrkh.mp4\n",
      "ctpqeykqdp.mp4\n",
      "cttqtsjvgn.mp4\n",
      "ctzmavwror.mp4\n",
      "curpwogllm.mp4\n",
      "cuzrgrbvil.mp4\n",
      "cvaksbpssm.mp4\n",
      "cwbacdwrzo.mp4\n",
      "cwqlvzefpg.mp4\n",
      "cwrtyzndpx.mp4\n",
      "cwsbspfzck.mp4\n",
      "cwwandrkus.mp4\n",
      "cxfujlvsuw.mp4\n",
      "cxrfacemmq.mp4\n",
      "cxttmymlbn.mp4\n",
      "cyboodqqyr.mp4\n",
      "cycacemkmt.mp4\n",
      "cyclgfjdrv.mp4\n",
      "czfunozvwp.mp4\n",
      "czkdanyadc.mp4\n",
      "czmqpxrqoh.mp4\n",
      "dafhtipaml.mp4\n",
      "dakqwktlbi.mp4\n",
      "dbhoxkblzx.mp4\n",
      "dbhrpizyeq.mp4\n",
      "dboxtiehng.mp4\n",
      "dbzcqmxzaj.mp4\n",
      "dbzpcjntve.mp4\n",
      "dcamvmuors.mp4\n",
      "dcuiiorugd.mp4\n",
      "ddhfabwpuz.mp4\n",
      "ddjggcasdw.mp4\n",
      "ddpvuimigj.mp4\n",
      "ddqccgmtka.mp4\n",
      "degpbqvcay.mp4\n",
      "deywhkarol.mp4\n",
      "deyyistcrd.mp4\n",
      "dfbpceeaox.mp4\n",
      "dgmevclvzy.mp4\n",
      "dgxrqjdomn.mp4\n",
      "dgzklxjmix.mp4\n",
      "dhcselezer.mp4\n",
      "dhevettufk.mp4\n",
      "dhjmzhrcav.mp4\n",
      "dhkwmjxwrn.mp4\n",
      "dhoqofwoxa.mp4\n",
      "diomeixhrg.mp4\n",
      "diopzaywor.mp4\n",
      "diqraixiov.mp4\n",
      "diuzrpqjli.mp4\n",
      "djvtbgwdcc.mp4\n",
      "djvutyvaio.mp4\n",
      "dkdwxmtpuo.mp4\n",
      "dkhlttuvmx.mp4\n",
      "dkrvorliqc.mp4\n",
      "dkwjwbwgey.mp4\n",
      "dlrsbscitn.mp4\n",
      "dnexlwbcxq.mp4\n",
      "dnhvalzvrt.mp4\n",
      "dntkzzzcdh.mp4\n",
      "dnyvfblxpm.mp4\n",
      "doanjploai.mp4\n",
      "dofusvhnib.mp4\n",
      "dozyddhild.mp4\n",
      "dptbnjnkdg.mp4\n",
      "dptrzdvwpg.mp4\n",
      "dqnyszdong.mp4\n",
      "dqppxmoqdl.mp4\n",
      "dqqtjcryjv.mp4\n"
     ]
    },
    {
     "ename": "ResourceExhaustedError",
     "evalue": " OOM when allocating tensor with shape[1,1150,646,10] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc\n\t [[{{node p_re_lu_38773/Relu_1-0-0-TransposeNCHWToNHWC-LayoutOptimizer}}]]\nHint: If you want to see a list of allocated tensors when OOM happens, add report_tensor_allocations_upon_oom to RunOptions for current allocation info.\n [Op:__inference_keras_scratch_graph_3695396]\n\nFunction call stack:\nkeras_scratch_graph\n",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mResourceExhaustedError\u001b[0m                    Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-245-a2ba00e1bdd3>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      3\u001b[0m     \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfake_video_file_name\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m     \u001b[0mvideo\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mVideo\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mos\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvideos_folder\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfake_video_file_name\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 5\u001b[1;33m     \u001b[0mfaces_from_video\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mextract_faces_from_video\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvideo\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m25\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmax_frames\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      6\u001b[0m     \u001b[0msave_crops\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'../to_be_annotated/'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfake_video_file_name\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'.'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfaces_from_video\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      7\u001b[0m     \u001b[0mremove_img_from_faces_from_video\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfaces_from_video\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-243-e731e3549ab1>\u001b[0m in \u001b[0;36mextract_faces_from_video\u001b[1;34m(video, frame_interval, max_frames)\u001b[0m\n\u001b[0;32m      6\u001b[0m     \u001b[1;32mfor\u001b[0m \u001b[0mframe_nr\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mstop\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mframe_interval\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      7\u001b[0m         \u001b[0mimg\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mvideo\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mframe_nr\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 8\u001b[1;33m         \u001b[0mfaces\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mextract_faces_from_img\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mframe_nr\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      9\u001b[0m         \u001b[0mfaces_from_video\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mupdate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfaces\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     10\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-243-e731e3549ab1>\u001b[0m in \u001b[0;36mextract_faces_from_img\u001b[1;34m(img, frame_nr)\u001b[0m\n\u001b[0;32m     49\u001b[0m     \u001b[0mfaces\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     50\u001b[0m     \u001b[0mdetector\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mMTCNN\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 51\u001b[1;33m     \u001b[0mresults\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdetector\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdetect_faces\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     52\u001b[0m     \u001b[1;32mfor\u001b[0m \u001b[0midx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mres\u001b[0m \u001b[1;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mresults\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     53\u001b[0m         \u001b[0mbox\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0marray\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mres\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'box'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\mtcnn\\mtcnn.py\u001b[0m in \u001b[0;36mdetect_faces\u001b[1;34m(self, img)\u001b[0m\n\u001b[0;32m    300\u001b[0m         \u001b[1;31m# We pipe here each of the stages\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    301\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mstage\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mstages\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 302\u001b[1;33m             \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mstage\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    303\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    304\u001b[0m         \u001b[1;33m[\u001b[0m\u001b[0mtotal_boxes\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpoints\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mresult\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\mtcnn\\mtcnn.py\u001b[0m in \u001b[0;36m__stage1\u001b[1;34m(self, image, scales, stage_status)\u001b[0m\n\u001b[0;32m    340\u001b[0m             \u001b[0mimg_y\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtranspose\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg_x\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m3\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    341\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 342\u001b[1;33m             \u001b[0mout\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_pnet\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mimg_y\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    343\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    344\u001b[0m             \u001b[0mout0\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtranspose\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mout\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m(\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m3\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mpredict\u001b[1;34m(self, x, batch_size, verbose, steps, callbacks, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[0;32m   1460\u001b[0m                                             \u001b[0mverbose\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mverbose\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1461\u001b[0m                                             \u001b[0msteps\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0msteps\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1462\u001b[1;33m                                             callbacks=callbacks)\n\u001b[0m\u001b[0;32m   1463\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1464\u001b[0m     def train_on_batch(self, x, y,\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\keras\\engine\\training_arrays.py\u001b[0m in \u001b[0;36mpredict_loop\u001b[1;34m(model, f, ins, batch_size, verbose, steps, callbacks)\u001b[0m\n\u001b[0;32m    322\u001b[0m             \u001b[0mbatch_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;34m'batch'\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mbatch_index\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'size'\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch_ids\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    323\u001b[0m             \u001b[0mcallbacks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_batch_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'predict'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m'begin'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch_index\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mbatch_logs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 324\u001b[1;33m             \u001b[0mbatch_outs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    325\u001b[0m             \u001b[0mbatch_outs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mto_list\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch_outs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    326\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mbatch_index\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\tensorflow_core\\python\\keras\\backend.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, inputs)\u001b[0m\n\u001b[0;32m   3738\u001b[0m         \u001b[0mvalue\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmath_ops\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcast\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtensor\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3739\u001b[0m       \u001b[0mconverted_inputs\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 3740\u001b[1;33m     \u001b[0moutputs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_graph_fn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0mconverted_inputs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   3741\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   3742\u001b[0m     \u001b[1;31m# EagerTensor.numpy() will often make a copy to ensure memory safety.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\tensorflow_core\\python\\eager\\function.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1079\u001b[0m       \u001b[0mTypeError\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mFor\u001b[0m \u001b[0minvalid\u001b[0m \u001b[0mpositional\u001b[0m\u001b[1;33m/\u001b[0m\u001b[0mkeyword\u001b[0m \u001b[0margument\u001b[0m \u001b[0mcombinations\u001b[0m\u001b[1;33m.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1080\u001b[0m     \"\"\"\n\u001b[1;32m-> 1081\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_impl\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1082\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1083\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_call_impl\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcancellation_manager\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\tensorflow_core\\python\\eager\\function.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, args, kwargs, cancellation_manager)\u001b[0m\n\u001b[0;32m   1119\u001b[0m       raise TypeError(\"Keyword arguments {} unknown. Expected {}.\".format(\n\u001b[0;32m   1120\u001b[0m           list(kwargs.keys()), list(self._arg_keywords)))\n\u001b[1;32m-> 1121\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call_flat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcaptured_inputs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcancellation_manager\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1122\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1123\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_filtered_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\tensorflow_core\\python\\eager\\function.py\u001b[0m in \u001b[0;36m_call_flat\u001b[1;34m(self, args, captured_inputs, cancellation_manager)\u001b[0m\n\u001b[0;32m   1222\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mexecuting_eagerly\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1223\u001b[0m       flat_outputs = forward_function.call(\n\u001b[1;32m-> 1224\u001b[1;33m           ctx, args, cancellation_manager=cancellation_manager)\n\u001b[0m\u001b[0;32m   1225\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1226\u001b[0m       \u001b[0mgradient_name\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_delayed_rewrite_functions\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mregister\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\tensorflow_core\\python\\eager\\function.py\u001b[0m in \u001b[0;36mcall\u001b[1;34m(self, ctx, args, cancellation_manager)\u001b[0m\n\u001b[0;32m    509\u001b[0m               \u001b[0minputs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    510\u001b[0m               \u001b[0mattrs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"executor_type\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mexecutor_type\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"config_proto\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mconfig\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 511\u001b[1;33m               ctx=ctx)\n\u001b[0m\u001b[0;32m    512\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    513\u001b[0m           outputs = execute.execute_with_cancellation(\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\tensorflow_core\\python\\eager\\execute.py\u001b[0m in \u001b[0;36mquick_execute\u001b[1;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[0;32m     65\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     66\u001b[0m       \u001b[0mmessage\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmessage\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 67\u001b[1;33m     \u001b[0msix\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mraise_from\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcore\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_status_to_exception\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0me\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcode\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmessage\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     68\u001b[0m   \u001b[1;32mexcept\u001b[0m \u001b[0mTypeError\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     69\u001b[0m     keras_symbolic_tensors = [\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\ml\\lib\\site-packages\\six.py\u001b[0m in \u001b[0;36mraise_from\u001b[1;34m(value, from_value)\u001b[0m\n",
      "\u001b[1;31mResourceExhaustedError\u001b[0m:  OOM when allocating tensor with shape[1,1150,646,10] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc\n\t [[{{node p_re_lu_38773/Relu_1-0-0-TransposeNCHWToNHWC-LayoutOptimizer}}]]\nHint: If you want to see a list of allocated tensors when OOM happens, add report_tensor_allocations_upon_oom to RunOptions for current allocation info.\n [Op:__inference_keras_scratch_graph_3695396]\n\nFunction call stack:\nkeras_scratch_graph\n"
     ]
    }
   ],
   "source": [
    "crops = {}\n",
    "for fake_video_file_name in df[df['label'] == 'FAKE']['index'].to_numpy():\n",
    "    print(fake_video_file_name)\n",
    "    video = Video(os.path.join(videos_folder, fake_video_file_name))\n",
    "    faces_from_video = extract_faces_from_video(video, 25, max_frames=20)\n",
    "    save_crops('../to_be_annotated_fake/', fake_video_file_name.split('.')[0], faces_from_video)\n",
    "    remove_img_from_faces_from_video(faces_from_video)\n",
    "    crops[fake_video_file_name] = faces_from_video\n",
    "save_obj(crops, 'crops')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T14:31:39.661567Z",
     "start_time": "2019-12-26T14:06:38.085655Z"
    },
    "deletable": false,
    "editable": false,
    "hidden": true,
    "run_control": {
     "frozen": true
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "abarnvbtwb.mp4\n",
      "aelfnikyqj.mp4\n",
      "afoovlsmtx.mp4\n",
      "agrmhtjdlk.mp4\n",
      "ahqqqilsxt.mp4\n",
      "ajqslcypsw.mp4\n",
      "anpuvshzoo.mp4\n",
      "asaxgevnnp.mp4\n",
      "atkdltyyen.mp4\n",
      "atvmxvwyns.mp4\n",
      "avmjormvsx.mp4\n",
      "axntxmycwd.mp4\n",
      "aybgughjxh.mp4\n",
      "aybumesmpk.mp4\n",
      "aytzyidmgs.mp4\n",
      "bddjdhzfze.mp4\n",
      "bdnaqemxmr.mp4\n",
      "beboztfcme.mp4\n",
      "bejhvclboh.mp4\n",
      "beyebyhrph.mp4\n",
      "bffwsjxghk.mp4\n",
      "bgvhtpzknn.mp4\n",
      "bgwmmujlmc.mp4\n",
      "bilnggbxgu.mp4\n",
      "bmjzrlszhi.mp4\n",
      "bpapbctoao.mp4\n",
      "brwrlczjvi.mp4\n",
      "bulkxhhknf.mp4\n",
      "bwhlgysghg.mp4\n",
      "bwipwzzxxu.mp4\n",
      "bxzakyopjf.mp4\n",
      "bzythlfnhq.mp4\n",
      "caifxvsozs.mp4\n",
      "ccfoszqabv.mp4\n",
      "cfxkpiweqt.mp4\n",
      "chtapglbcj.mp4\n",
      "chviwxsfhg.mp4\n",
      "ciyoudyhly.mp4\n",
      "cizlkenljw.mp4\n",
      "ckjaibzfxa.mp4\n",
      "ckkuyewywx.mp4\n",
      "clrycekyst.mp4\n",
      "cmbzllswnl.mp4\n",
      "cobjrlugvp.mp4\n",
      "cpjxareypw.mp4\n",
      "cppdvdejkc.mp4\n",
      "cprhtltsjp.mp4\n",
      "crezycjqyk.mp4\n",
      "cyxlcuyznd.mp4\n",
      "dakiztgtnw.mp4\n",
      "dbnygxtwek.mp4\n",
      "dbtbbhakdv.mp4\n",
      "ddepeddixj.mp4\n",
      "dhcndnuwta.mp4\n",
      "dhxctgyoqj.mp4\n",
      "djxdyjopjd.mp4\n",
      "dkuayagnmc.mp4\n",
      "dkzvdrzcnr.mp4\n",
      "dlpoieqvfb.mp4\n",
      "drcyabprvt.mp4\n",
      "dsjbknkujw.mp4\n",
      "duycddgtrl.mp4\n",
      "dxbqjxrhin.mp4\n",
      "dzyuwjkjui.mp4\n",
      "eckvhdusax.mp4\n",
      "ecujsjhscd.mp4\n",
      "edyncaijwx.mp4\n",
      "efwfxwwlbw.mp4\n",
      "eggbjzxnmg.mp4\n",
      "egghxjjmfg.mp4\n",
      "ehccixxzoe.mp4\n",
      "ehtdtkmmli.mp4\n",
      "ekcrtigpab.mp4\n",
      "ellavthztb.mp4\n",
      "eqnoqyfquo.mp4\n",
      "erlvuvjsjf.mp4\n",
      "eudeqjhdfd.mp4\n"
     ]
    }
   ],
   "source": [
    "crops = {}\n",
    "for fake_video_file_name in df[df['label'] == 'REAL']['index'].to_numpy():\n",
    "    print(fake_video_file_name)\n",
    "    video = Video(os.path.join(videos_folder, fake_video_file_name))\n",
    "    faces_from_video = extract_faces_from_video(video, 25, max_frames=20)\n",
    "    save_crops('../to_be_annotated_real/', fake_video_file_name.split('.')[0], faces_from_video)\n",
    "    remove_img_from_faces_from_video(faces_from_video)\n",
    "    crops[fake_video_file_name] = faces_from_video\n",
    "save_obj(crops, 'crops_real', fake=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:10:37.813648Z",
     "start_time": "2019-12-26T15:10:37.786648Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "crops_real = load_obj('crops_real', fake=False)\n",
    "crops_fake = load_obj('crops_fake', fake=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:10:41.680260Z",
     "start_time": "2019-12-26T15:10:41.377259Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'abarnvbtwb.mp4': {'0_0': {'coordinates': {'x1': 789,\n",
       "    'y1': 198,\n",
       "    'width': 206,\n",
       "    'height': 211},\n",
       "   'keypoints': {'left_eye': (847, 301),\n",
       "    'right_eye': (918, 248),\n",
       "    'nose': (908, 294),\n",
       "    'mouth_left': (903, 359),\n",
       "    'mouth_right': (962, 315)},\n",
       "   'confidence': 0.9996341466903687},\n",
       "  '25_0': {'coordinates': {'x1': 852, 'y1': 162, 'width': 198, 'height': 232},\n",
       "   'keypoints': {'left_eye': (911, 264),\n",
       "    'right_eye': (996, 240),\n",
       "    'nose': (973, 279),\n",
       "    'mouth_left': (940, 339),\n",
       "    'mouth_right': (1022, 317)},\n",
       "   'confidence': 0.9999998807907104},\n",
       "  '50_0': {'coordinates': {'x1': 939, 'y1': 156, 'width': 177, 'height': 243},\n",
       "   'keypoints': {'left_eye': (1033, 249),\n",
       "    'right_eye': (1108, 264),\n",
       "    'nose': (1087, 294),\n",
       "    'mouth_left': (1017, 325),\n",
       "    'mouth_right': (1095, 341)},\n",
       "   'confidence': 0.9999992847442627},\n",
       "  '75_0': {'coordinates': {'x1': 906, 'y1': 152, 'width': 185, 'height': 241},\n",
       "   'keypoints': {'left_eye': (972, 248),\n",
       "    'right_eye': (1057, 250),\n",
       "    'nose': (1022, 289),\n",
       "    'mouth_left': (967, 325),\n",
       "    'mouth_right': (1055, 329)},\n",
       "   'confidence': 1.0},\n",
       "  '100_0': {'coordinates': {'x1': 899, 'y1': 147, 'width': 184, 'height': 236},\n",
       "   'keypoints': {'left_eye': (958, 239),\n",
       "    'right_eye': (1043, 235),\n",
       "    'nose': (1004, 275),\n",
       "    'mouth_left': (958, 316),\n",
       "    'mouth_right': (1048, 315)},\n",
       "   'confidence': 0.9999920129776001},\n",
       "  '125_0': {'coordinates': {'x1': 900, 'y1': 148, 'width': 179, 'height': 225},\n",
       "   'keypoints': {'left_eye': (960, 233),\n",
       "    'right_eye': (1044, 230),\n",
       "    'nose': (1010, 263),\n",
       "    'mouth_left': (973, 316),\n",
       "    'mouth_right': (1041, 314)},\n",
       "   'confidence': 0.9999997615814209},\n",
       "  '150_0': {'coordinates': {'x1': 895, 'y1': 184, 'width': 194, 'height': 264},\n",
       "   'keypoints': {'left_eye': (961, 293),\n",
       "    'right_eye': (1047, 290),\n",
       "    'nose': (1009, 338),\n",
       "    'mouth_left': (970, 385),\n",
       "    'mouth_right': (1042, 385)},\n",
       "   'confidence': 0.9999947547912598},\n",
       "  '175_0': {'coordinates': {'x1': 895, 'y1': 182, 'width': 183, 'height': 252},\n",
       "   'keypoints': {'left_eye': (952, 278),\n",
       "    'right_eye': (1034, 275),\n",
       "    'nose': (998, 323),\n",
       "    'mouth_left': (962, 373),\n",
       "    'mouth_right': (1030, 371)},\n",
       "   'confidence': 0.9999897480010986},\n",
       "  '200_0': {'coordinates': {'x1': 883, 'y1': 180, 'width': 186, 'height': 249},\n",
       "   'keypoints': {'left_eye': (949, 285),\n",
       "    'right_eye': (1031, 281),\n",
       "    'nose': (992, 327),\n",
       "    'mouth_left': (956, 375),\n",
       "    'mouth_right': (1024, 372)},\n",
       "   'confidence': 0.9999992847442627},\n",
       "  '225_0': {'coordinates': {'x1': 889, 'y1': 165, 'width': 188, 'height': 242},\n",
       "   'keypoints': {'left_eye': (951, 259),\n",
       "    'right_eye': (1037, 256),\n",
       "    'nose': (1001, 297),\n",
       "    'mouth_left': (951, 336),\n",
       "    'mouth_right': (1041, 335)},\n",
       "   'confidence': 0.9999998807907104},\n",
       "  '250_0': {'coordinates': {'x1': 912, 'y1': 166, 'width': 183, 'height': 252},\n",
       "   'keypoints': {'left_eye': (985, 261),\n",
       "    'right_eye': (1068, 277),\n",
       "    'nose': (1028, 312),\n",
       "    'mouth_left': (964, 338),\n",
       "    'mouth_right': (1057, 354)},\n",
       "   'confidence': 1.0},\n",
       "  '275_0': {'coordinates': {'x1': 950, 'y1': 159, 'width': 197, 'height': 252},\n",
       "   'keypoints': {'left_eye': (1030, 247),\n",
       "    'right_eye': (1112, 275),\n",
       "    'nose': (1064, 300),\n",
       "    'mouth_left': (1008, 339),\n",
       "    'mouth_right': (1073, 361)},\n",
       "   'confidence': 0.9999920129776001}},\n",
       " 'aelfnikyqj.mp4': {'0_0': {'coordinates': {'x1': 908,\n",
       "    'y1': 61,\n",
       "    'width': 95,\n",
       "    'height': 128},\n",
       "   'keypoints': {'left_eye': (933, 112),\n",
       "    'right_eye': (979, 109),\n",
       "    'nose': (958, 134),\n",
       "    'mouth_left': (941, 161),\n",
       "    'mouth_right': (976, 161)},\n",
       "   'confidence': 0.9999853372573853},\n",
       "  '25_0': {'coordinates': {'x1': 928, 'y1': 59, 'width': 93, 'height': 126},\n",
       "   'keypoints': {'left_eye': (954, 107),\n",
       "    'right_eye': (999, 105),\n",
       "    'nose': (981, 128),\n",
       "    'mouth_left': (961, 157),\n",
       "    'mouth_right': (999, 157)},\n",
       "   'confidence': 0.9999964237213135},\n",
       "  '50_0': {'coordinates': {'x1': 921, 'y1': 64, 'width': 94, 'height': 127},\n",
       "   'keypoints': {'left_eye': (947, 113),\n",
       "    'right_eye': (992, 109),\n",
       "    'nose': (973, 134),\n",
       "    'mouth_left': (955, 162),\n",
       "    'mouth_right': (993, 160)},\n",
       "   'confidence': 0.9999946355819702},\n",
       "  '75_0': {'coordinates': {'x1': 916, 'y1': 61, 'width': 97, 'height': 131},\n",
       "   'keypoints': {'left_eye': (942, 112),\n",
       "    'right_eye': (988, 108),\n",
       "    'nose': (967, 133),\n",
       "    'mouth_left': (951, 162),\n",
       "    'mouth_right': (987, 161)},\n",
       "   'confidence': 0.9999858140945435},\n",
       "  '100_0': {'coordinates': {'x1': 914, 'y1': 47, 'width': 100, 'height': 135},\n",
       "   'keypoints': {'left_eye': (946, 99),\n",
       "    'right_eye': (992, 100),\n",
       "    'nose': (975, 122),\n",
       "    'mouth_left': (953, 151),\n",
       "    'mouth_right': (989, 152)},\n",
       "   'confidence': 0.9999979734420776},\n",
       "  '125_0': {'coordinates': {'x1': 911, 'y1': 53, 'width': 100, 'height': 135},\n",
       "   'keypoints': {'left_eye': (941, 104),\n",
       "    'right_eye': (988, 104),\n",
       "    'nose': (969, 127),\n",
       "    'mouth_left': (948, 157),\n",
       "    'mouth_right': (984, 158)},\n",
       "   'confidence': 0.9999873638153076},\n",
       "  '150_0': {'coordinates': {'x1': 916, 'y1': 59, 'width': 97, 'height': 134},\n",
       "   'keypoints': {'left_eye': (947, 108),\n",
       "    'right_eye': (991, 110),\n",
       "    'nose': (973, 130),\n",
       "    'mouth_left': (949, 159),\n",
       "    'mouth_right': (988, 161)},\n",
       "   'confidence': 0.9999971389770508},\n",
       "  '175_0': {'coordinates': {'x1': 904, 'y1': 60, 'width': 95, 'height': 132},\n",
       "   'keypoints': {'left_eye': (934, 110),\n",
       "    'right_eye': (979, 108),\n",
       "    'nose': (961, 132),\n",
       "    'mouth_left': (938, 160),\n",
       "    'mouth_right': (977, 159)},\n",
       "   'confidence': 0.9998842477798462},\n",
       "  '200_0': {'coordinates': {'x1': 902, 'y1': 54, 'width': 95, 'height': 133},\n",
       "   'keypoints': {'left_eye': (924, 108),\n",
       "    'right_eye': (969, 101),\n",
       "    'nose': (948, 128),\n",
       "    'mouth_left': (933, 159),\n",
       "    'mouth_right': (972, 154)},\n",
       "   'confidence': 0.9999864101409912},\n",
       "  '225_0': {'coordinates': {'x1': 900, 'y1': 62, 'width': 94, 'height': 129},\n",
       "   'keypoints': {'left_eye': (924, 112),\n",
       "    'right_eye': (969, 106),\n",
       "    'nose': (949, 131),\n",
       "    'mouth_left': (935, 163),\n",
       "    'mouth_right': (970, 159)},\n",
       "   'confidence': 0.9999905824661255},\n",
       "  '250_0': {'coordinates': {'x1': 901, 'y1': 64, 'width': 97, 'height': 129},\n",
       "   'keypoints': {'left_eye': (928, 116),\n",
       "    'right_eye': (973, 110),\n",
       "    'nose': (954, 136),\n",
       "    'mouth_left': (939, 165),\n",
       "    'mouth_right': (976, 161)},\n",
       "   'confidence': 0.9999939203262329},\n",
       "  '250_1': {'coordinates': {'x1': 468, 'y1': 640, 'width': 27, 'height': 30},\n",
       "   'keypoints': {'left_eye': (472, 648),\n",
       "    'right_eye': (485, 648),\n",
       "    'nose': (473, 654),\n",
       "    'mouth_left': (471, 663),\n",
       "    'mouth_right': (480, 663)},\n",
       "   'confidence': 0.9621386528015137},\n",
       "  '275_0': {'coordinates': {'x1': 912, 'y1': 59, 'width': 98, 'height': 138},\n",
       "   'keypoints': {'left_eye': (936, 111),\n",
       "    'right_eye': (981, 108),\n",
       "    'nose': (960, 130),\n",
       "    'mouth_left': (943, 163),\n",
       "    'mouth_right': (981, 161)},\n",
       "   'confidence': 0.9999946355819702},\n",
       "  '275_1': {'coordinates': {'x1': 468, 'y1': 640, 'width': 26, 'height': 28},\n",
       "   'keypoints': {'left_eye': (472, 649),\n",
       "    'right_eye': (484, 648),\n",
       "    'nose': (473, 654),\n",
       "    'mouth_left': (472, 663),\n",
       "    'mouth_right': (481, 663)},\n",
       "   'confidence': 0.9642264246940613}},\n",
       " 'afoovlsmtx.mp4': {'0_0': {'coordinates': {'x1': 328,\n",
       "    'y1': 675,\n",
       "    'width': 251,\n",
       "    'height': 329},\n",
       "   'keypoints': {'left_eye': (381, 800),\n",
       "    'right_eye': (492, 799),\n",
       "    'nose': (425, 831),\n",
       "    'mouth_left': (386, 917),\n",
       "    'mouth_right': (482, 918)},\n",
       "   'confidence': 0.9997612833976746},\n",
       "  '25_0': {'coordinates': {'x1': 331, 'y1': 678, 'width': 234, 'height': 318},\n",
       "   'keypoints': {'left_eye': (382, 796),\n",
       "    'right_eye': (490, 795),\n",
       "    'nose': (432, 830),\n",
       "    'mouth_left': (393, 905),\n",
       "    'mouth_right': (482, 902)},\n",
       "   'confidence': 0.9999977350234985},\n",
       "  '50_0': {'coordinates': {'x1': 321, 'y1': 687, 'width': 250, 'height': 324},\n",
       "   'keypoints': {'left_eye': (380, 815),\n",
       "    'right_eye': (488, 814),\n",
       "    'nose': (425, 854),\n",
       "    'mouth_left': (392, 929),\n",
       "    'mouth_right': (485, 931)},\n",
       "   'confidence': 0.9999864101409912},\n",
       "  '75_0': {'coordinates': {'x1': 347, 'y1': 705, 'width': 222, 'height': 290},\n",
       "   'keypoints': {'left_eye': (414, 814),\n",
       "    'right_eye': (516, 818),\n",
       "    'nose': (473, 858),\n",
       "    'mouth_left': (429, 922),\n",
       "    'mouth_right': (503, 926)},\n",
       "   'confidence': 0.9999949932098389},\n",
       "  '100_0': {'coordinates': {'x1': 331, 'y1': 700, 'width': 225, 'height': 296},\n",
       "   'keypoints': {'left_eye': (393, 823),\n",
       "    'right_eye': (500, 825),\n",
       "    'nose': (447, 862),\n",
       "    'mouth_left': (396, 930),\n",
       "    'mouth_right': (488, 935)},\n",
       "   'confidence': 0.9999840259552002},\n",
       "  '125_0': {'coordinates': {'x1': 327, 'y1': 692, 'width': 226, 'height': 291},\n",
       "   'keypoints': {'left_eye': (387, 815),\n",
       "    'right_eye': (486, 807),\n",
       "    'nose': (442, 859),\n",
       "    'mouth_left': (407, 921),\n",
       "    'mouth_right': (488, 913)},\n",
       "   'confidence': 0.9999496936798096},\n",
       "  '150_0': {'coordinates': {'x1': 316, 'y1': 699, 'width': 242, 'height': 316},\n",
       "   'keypoints': {'left_eye': (386, 826),\n",
       "    'right_eye': (496, 826),\n",
       "    'nose': (437, 863),\n",
       "    'mouth_left': (401, 934),\n",
       "    'mouth_right': (493, 939)},\n",
       "   'confidence': 0.9999878406524658},\n",
       "  '175_0': {'coordinates': {'x1': 287, 'y1': 697, 'width': 222, 'height': 300},\n",
       "   'keypoints': {'left_eye': (337, 822),\n",
       "    'right_eye': (438, 817),\n",
       "    'nose': (382, 866),\n",
       "    'mouth_left': (352, 932),\n",
       "    'mouth_right': (431, 925)},\n",
       "   'confidence': 0.9999754428863525},\n",
       "  '200_0': {'coordinates': {'x1': 312, 'y1': 699, 'width': 240, 'height': 308},\n",
       "   'keypoints': {'left_eye': (378, 820),\n",
       "    'right_eye': (486, 824),\n",
       "    'nose': (428, 854),\n",
       "    'mouth_left': (383, 931),\n",
       "    'mouth_right': (481, 936)},\n",
       "   'confidence': 0.9999862909317017},\n",
       "  '225_0': {'coordinates': {'x1': 318, 'y1': 689, 'width': 240, 'height': 295},\n",
       "   'keypoints': {'left_eye': (380, 810),\n",
       "    'right_eye': (487, 822),\n",
       "    'nose': (417, 859),\n",
       "    'mouth_left': (367, 909),\n",
       "    'mouth_right': (473, 918)},\n",
       "   'confidence': 0.999998927116394},\n",
       "  '250_0': {'coordinates': {'x1': 332, 'y1': 744, 'width': 221, 'height': 265},\n",
       "   '"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<b>limit_output extension: Maximum message size of 10000 exceeded with 297186 characters</b>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "crops_real"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:10:45.954260Z",
     "start_time": "2019-12-26T15:10:45.084283Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'aagfhgtpmv.mp4': {'0_0': {'coordinates': {'x1': 338,\n",
       "    'y1': 167,\n",
       "    'width': 219,\n",
       "    'height': 291},\n",
       "   'keypoints': {'left_eye': (416, 279),\n",
       "    'right_eye': (514, 266),\n",
       "    'nose': (495, 316),\n",
       "    'mouth_left': (452, 393),\n",
       "    'mouth_right': (527, 384)},\n",
       "   'confidence': 0.9999223947525024},\n",
       "  '25_0': {'coordinates': {'x1': 503, 'y1': 173, 'width': 220, 'height': 286},\n",
       "   'keypoints': {'left_eye': (588, 267),\n",
       "    'right_eye': (690, 287),\n",
       "    'nose': (648, 316),\n",
       "    'mouth_left': (588, 382),\n",
       "    'mouth_right': (667, 399)},\n",
       "   'confidence': 0.9999939203262329},\n",
       "  '50_0': {'coordinates': {'x1': 458, 'y1': 177, 'width': 219, 'height': 290},\n",
       "   'keypoints': {'left_eye': (520, 276),\n",
       "    'right_eye': (624, 288),\n",
       "    'nose': (556, 321),\n",
       "    'mouth_left': (508, 389),\n",
       "    'mouth_right': (602, 401)},\n",
       "   'confidence': 0.9999035596847534},\n",
       "  '75_0': {'coordinates': {'x1': 851, 'y1': 202, 'width': 198, 'height': 267},\n",
       "   'keypoints': {'left_eye': (900, 306),\n",
       "    'right_eye': (992, 303),\n",
       "    'nose': (943, 349),\n",
       "    'mouth_left': (910, 412),\n",
       "    'mouth_right': (985, 410)},\n",
       "   'confidence': 0.9999959468841553},\n",
       "  '100_0': {'coordinates': {'x1': 1072,\n",
       "    'y1': 199,\n",
       "    'width': 204,\n",
       "    'height': 288},\n",
       "   'keypoints': {'left_eye': (1121, 300),\n",
       "    'right_eye': (1222, 314),\n",
       "    'nose': (1154, 350),\n",
       "    'mouth_left': (1106, 408),\n",
       "    'mouth_right': (1196, 424)},\n",
       "   'confidence': 0.9999788999557495},\n",
       "  '125_0': {'coordinates': {'x1': 1091,\n",
       "    'y1': 167,\n",
       "    'width': 206,\n",
       "    'height': 284},\n",
       "   'keypoints': {'left_eye': (1138, 278),\n",
       "    'right_eye': (1237, 273),\n",
       "    'nose': (1182, 319),\n",
       "    'mouth_left': (1147, 386),\n",
       "    'mouth_right': (1232, 384)},\n",
       "   'confidence': 0.9995755553245544},\n",
       "  '150_0': {'coordinates': {'x1': 1020,\n",
       "    'y1': 189,\n",
       "    'width': 192,\n",
       "    'height': 275},\n",
       "   'keypoints': {'left_eye': (1058, 292),\n",
       "    'right_eye': (1146, 294),\n",
       "    'nose': (1081, 339),\n",
       "    'mouth_left': (1056, 396),\n",
       "    'mouth_right': (1132, 402)},\n",
       "   'confidence': 0.9999614953994751},\n",
       "  '175_0': {'coordinates': {'x1': 1060,\n",
       "    'y1': 120,\n",
       "    'width': 206,\n",
       "    'height': 266},\n",
       "   'keypoints': {'left_eye': (1107, 222),\n",
       "    'right_eye': (1202, 230),\n",
       "    'nose': (1133, 263),\n",
       "    'mouth_left': (1092, 318),\n",
       "    'mouth_right': (1180, 329)},\n",
       "   'confidence': 0.9999535083770752},\n",
       "  '200_0': {'coordinates': {'x1': 925, 'y1': 110, 'width': 206, 'height': 269},\n",
       "   'keypoints': {'left_eye': (981, 220),\n",
       "    'right_eye': (1075, 207),\n",
       "    'nose': (1036, 257),\n",
       "    'mouth_left': (1005, 318),\n",
       "    'mouth_right': (1081, 310)},\n",
       "   'confidence': 0.999936580657959},\n",
       "  '225_0': {'coordinates': {'x1': 937, 'y1': 117, 'width': 199, 'height': 269},\n",
       "   'keypoints': {'left_eye': (992, 218),\n",
       "    'right_eye': (1085, 211),\n",
       "    'nose': (1049, 253),\n",
       "    'mouth_left': (1010, 318),\n",
       "    'mouth_right': (1090, 314)},\n",
       "   'confidence': 0.9998329877853394},\n",
       "  '250_0': {'coordinates': {'x1': 940, 'y1': 156, 'width': 183, 'height': 250},\n",
       "   'keypoints': {'left_eye': (1002, 251),\n",
       "    'right_eye': (1088, 250),\n",
       "    'nose': (1057, 295),\n",
       "    'mouth_left': (1018, 353),\n",
       "    'mouth_right': (1082, 353)},\n",
       "   'confidence': 0.9999988079071045},\n",
       "  '275_0': {'coordinates': {'x1': 618, 'y1': 142, 'width': 211, 'height': 281},\n",
       "   'keypoints': {'left_eye': (694, 252),\n",
       "    'right_eye': (790, 249),\n",
       "    'nose': (760, 299),\n",
       "    'mouth_left': (712, 359),\n",
       "    'mouth_right': (788, 357)},\n",
       "   'confidence': 0.9999850988388062}},\n",
       " 'aapnvogymq.mp4': {'0_0': {'coordinates': {'x1': 1189,\n",
       "    'y1': 192,\n",
       "    'width': 146,\n",
       "    'height': 185},\n",
       "   'keypoints': {'left_eye': (1209, 262),\n",
       "    'right_eye': (1248, 255),\n",
       "    'nose': (1210, 300),\n",
       "    'mouth_left': (1231, 338),\n",
       "    'mouth_right': (1260, 333)},\n",
       "   'confidence': 0.9999994039535522},\n",
       "  '0_1': {'coordinates': {'x1': 585, 'y1': 241, 'width': 126, 'height': 191},\n",
       "   'keypoints': {'left_eye': (664, 318),\n",
       "    'right_eye': (697, 316),\n",
       "    'nose': (707, 350),\n",
       "    'mouth_left': (673, 389),\n",
       "    'mouth_right': (696, 387)},\n",
       "   'confidence': 0.9999772310256958},\n",
       "  '25_0': {'coordinates': {'x1': 1183, 'y1': 184, 'width': 153, 'height': 190},\n",
       "   'keypoints': {'left_eye': (1217, 258),\n",
       "    'right_eye': (1269, 246),\n",
       "    'nose': (1233, 289),\n",
       "    'mouth_left': (1239, 333),\n",
       "    'mouth_right': (1279, 325)},\n",
       "   'confidence': 0.9999915361404419},\n",
       "  '25_1': {'coordinates': {'x1': 554, 'y1': 234, 'width': 137, 'height': 215},\n",
       "   'keypoints': {'left_eye': (641, 315),\n",
       "    'right_eye': (677, 307),\n",
       "    'nose': (694, 349),\n",
       "    'mouth_left': (656, 399),\n",
       "    'mouth_right': (685, 393)},\n",
       "   'confidence': 0.9998846054077148},\n",
       "  '50_0': {'coordinates': {'x1': 1181, 'y1': 190, 'width': 144, 'height': 186},\n",
       "   'keypoints': {'left_eye': (1196, 264),\n",
       "    'right_eye': (1241, 257),\n",
       "    'nose': (1199, 301),\n",
       "    'mouth_left': (1217, 342),\n",
       "    'mouth_right': (1253, 335)},\n",
       "   'confidence': 0.9999994039535522},\n",
       "  '50_1': {'coordinates': {'x1': 557, 'y1': 224, 'width': 142, 'height': 206},\n",
       "   'keypoints': {'left_eye': (629, 300),\n",
       "    'right_eye': (686, 304),\n",
       "    'nose': (679, 338),\n",
       "    'mouth_left': (634, 380),\n",
       "    'mouth_right': (683, 381)},\n",
       "   'confidence': 0.999972939491272},\n",
       "  '75_0': {'coordinates': {'x1': 1179, 'y1': 186, 'width': 143, 'height': 188},\n",
       "   'keypoints': {'left_eye': (1193, 263),\n",
       "    'right_eye': (1237, 256),\n",
       "    'nose': (1197, 299),\n",
       "    'mouth_left': (1214, 341),\n",
       "    'mouth_right': (1248, 335)},\n",
       "   'confidence': 0.9999995231628418},\n",
       "  '75_1': {'coordinates': {'x1': 551, 'y1': 225, 'width': 144, 'height': 200},\n",
       "   'keypoints': {'left_eye': (615, 305),\n",
       "    'right_eye': (675, 301),\n",
       "    'nose': (666, 336),\n",
       "    'mouth_left': (626, 378),\n",
       "    'mouth_right': (677, 373)},\n",
       "   'confidence': 0.9999712705612183},\n",
       "  '100_0': {'coordinates': {'x1': 1174,\n",
       "    'y1': 185,\n",
       "    'width': 146,\n",
       "    'height': 192},\n",
       "   'keypoints': {'left_eye': (1188, 261),\n",
       "    'right_eye': (1236, 254),\n",
       "    'nose': (1193, 298),\n",
       "    'mouth_left': (1209, 341),\n",
       "    'mouth_right': (1247, 335)},\n",
       "   'confidence': 0.9999994039535522},\n",
       "  '100_1': {'coordinates': {'x1': 550, 'y1': 212, 'width': 153, 'height': 224},\n",
       "   'keypoints': {'left_eye': (622, 302),\n",
       "    'right_eye': (682, 299),\n",
       "    'nose': (675, 334),\n",
       "    'mouth_left': (627, 380),\n",
       "    'mouth_right': (681, 374)},\n",
       "   'confidence': 0.9999619722366333},\n",
       "  '125_0': {'coordinates': {'x1': 1173,\n",
       "    'y1': 187,\n",
       "    'width': 143,\n",
       "    'height': 187},\n",
       "   'keypoints': {'left_eye': (1186, 261),\n",
       "    'right_eye': (1234, 253),\n",
       "    'nose': (1192, 297),\n",
       "    'mouth_left': (1207, 340),\n",
       "    'mouth_right': (1244, 333)},\n",
       "   'confidence': 0.9999996423721313},\n",
       "  '125_1': {'coordinates': {'x1': 535, 'y1': 238, 'width': 141, 'height': 198},\n",
       "   'keypoints': {'left_eye': (613, 312),\n",
       "    'right_eye': (666, 314),\n",
       "    'nose': (664, 351),\n",
       "    'mouth_left': (622, 392),\n",
       "    'mouth_right': (666, 392)},\n",
       "   'confidence': 0.9987745881080627},\n",
       "  '150_0': {'coordinates': {'x1': 1168,\n",
       "    'y1': 187,\n",
       "    'width': 143,\n",
       "    'height': 186},\n",
       "   'keypoints': {'left_eye': (1183, 263),\n",
       "    'right_eye': (1231, 253),\n",
       "    'nose': (1191, 298),\n",
       "    'mouth_left': (1205, 341),\n",
       "    'mouth_right': (1243, 333)},\n",
       "   'confidence': 0.9999995231628418},\n",
       "  '150_1': {'coordinates': {'x1': 534, 'y1': 236, 'width': 142, 'height': 206},\n",
       "   'keypoints': {'left_eye': (612, 313),\n",
       "    'right_eye': (667, 319),\n",
       "    'nose': (662, 353),\n",
       "    'mouth_left': (612, 395),\n",
       "    'mouth_right': (660, 397)},\n",
       "   'confidence': 0.9990805387496948},\n",
       "  '175_0': {'coordinates': {'x1': 1158,\n",
       "    'y1': 185,\n",
       "    'width': 145,\n",
       "    'height': 188},\n",
       "   'keypoints': {'left_eye': (1175, 262),\n",
       "    'right_eye': (1226, 253),\n",
       "    'nose': (1185, 299),\n",
       "    'mouth_left': (1197, 341),\n",
       "    'mouth_right': (1234, 333)},\n",
       "   'confidence': 0.9999995231628418},\n",
       "  '175_1': {'coordinates': {'x1': 563, 'y1': 219, 'width': 141, 'height': 237},\n",
       "   'keypoints': {'left_eye': (654, 308),\n",
       "    'right_eye': (686, 306),\n",
       "    'nose': (699, 350),\n",
       "    'mouth_left': (658, 397),\n",
       "    'mouth_right': (680, 396)},\n",
       "   'confidence': 0.9999889135360718},\n",
       "  '200_0': {'coordinates': {'x1': 1161,\n",
       "    'y1': 189,\n",
       "    'width': 158,\n",
       "    'height': 187},\n",
       "   'keypoints': {'left_eye': (1191, 274),\n",
       "    'right_eye': (1255, 262),\n",
       "    'nose': (1218, 310),\n",
       "    'mouth_left': (1212, 345),\n",
       "    'mouth_right': (1268, 333)},\n",
       "   'confidence': 0.9999996423721313},\n",
       "  '200_1': {'coordinates': {'x1': 545, 'y1': 225, 'width': 135, 'height': 223},\n",
       "   'keypoints': {'left_eye': (633, 306),\n",
       "    'right_eye': (672, 308),\n",
       "    'nose': (680, 345),\n",
       "    'mouth_left': (631, 388),\n",
       "    'mouth_right': (666, 390)},\n",
       "   'confidence': 0.9999912977218628},\n",
       "  '225_0': {'coordinates': {'x1': 1157,\n",
       "    'y1': 184,\n",
       "    'width': 141,\n",
       "    'height': 186},\n",
       "   'keypoints': {'left_eye': (1169, 256),\n",
       "    'right_eye': (1219, 251),\n",
       "    'nose': (1174, 292),\n",
       "    'mouth_left': (1187, 334),\n",
       "    'mouth_right': (1225, 331)},\n",
       "   'confidence': 0.9999996423721313},\n",
       "  '225_1': {'coordinates': {'x1': 526, 'y1': 220, 'width': 149, 'height': 211},\n",
       "   'keypoints': {'left_eye': (591, 301),\n",
       "    'right_eye': (655, 304),\n",
       "    'nose': (638, 336),\n",
       "    'mouth_left': (593, 375),\n",
       "    'mouth_right': (653, 376)},\n",
       "   'confidence': 0.9999985694885254},\n",
       "  '250_0': {'coordinates': {'x1': 1172,\n",
       "    'y1': 180,\n",
       "    'width': 142,\n",
       "    'height': 189},\n",
       "   'keypoints': {'left_eye': (1185, 254),\n",
       "    'right_eye': (1234, 248),\n",
       "    'nose': (1191, 292),\n",
       "    'mouth_left': (1202, 334),\n",
       "    'mouth_right': (1240, 330)},\n",
       "   'confidence': 0.9999998807907104},\n",
       "  '250_1': {'coordinates': {'x1': 535, 'y1': 227, 'width': 144, 'height': 200},\n",
       "   'keypoints': {'left_eye': (605, 305),\n",
       "    'right_eye': (663, 302),\n",
       "    'nose': (653, 343),\n",
       "    'mouth_left': (611, 383),\n",
       "    'mouth_right': (660, 381)},\n",
       "   'confidence': 0.9990209341049194},\n",
       "  '275_0': {'coordinates': {'x1': 1179,\n",
       "    'y1': 177,\n",
       "    'width': 146,\n",
       "    'height': 195},\n",
       "   'keypoints': {'left_eye': (1193, 252),\n",
       "    'right_eye': (1244, 248),\n",
       "    'nose': (1198, 292),\n",
       "    'mouth_left': (1210, 334),\n",
       "    'mouth_right': (1248, 332)},\n",
       "   'confidence': 0.9999992847442627},\n",
       "  '275_1': {'coordinates': {'x1': 535, 'y1': 226, 'width': 140, 'height': 213},\n",
       "   'keypoints': {'left_eye': (611"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<b>limit_output extension: Maximum message size of 10000 exceeded with 829018 characters</b>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "crops_fake"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:15:53.405837Z",
     "start_time": "2019-12-26T15:15:53.402834Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "keys = [x for x in crops_fake['aagfhgtpmv.mp4'].keys() if x.split('_')[1] == '0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:19:48.467395Z",
     "start_time": "2019-12-26T15:19:48.463421Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "d = {key: crops_fake['aagfhgtpmv.mp4'][key] for key in keys}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:19:49.041911Z",
     "start_time": "2019-12-26T15:19:49.037909Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['0_0', '25_0', '50_0', '75_0', '100_0', '125_0', '150_0', '175_0', '200_0', '225_0', '250_0', '275_0'])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:20:55.781788Z",
     "start_time": "2019-12-26T15:20:55.778815Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "obj = d['0_0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:23:50.931433Z",
     "start_time": "2019-12-26T15:23:50.927431Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'coordinates': {'x1': 338, 'y1': 167, 'width': 219, 'height': 291},\n",
       " 'keypoints': {'left_eye': (416, 279),\n",
       "  'right_eye': (514, 266),\n",
       "  'nose': (495, 316),\n",
       "  'mouth_left': (452, 393),\n",
       "  'mouth_right': (527, 384)},\n",
       " 'confidence': 0.9999223947525024}"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "obj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:24:18.282945Z",
     "start_time": "2019-12-26T15:24:18.279923Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "obj2 = d['25_0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:32:53.621674Z",
     "start_time": "2019-12-26T15:32:53.613675Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "def sum_distances(keypoints):\n",
    "    k = keypoints\n",
    "    return np.array([np.sum(np.abs(np.array(k['left_eye']) - np.array(k['right_eye']))),\n",
    "                  np.sum(np.abs(np.array(k['left_eye']) - np.array(k['nose']))),\n",
    "                  np.sum(np.abs(np.array(k['right_eye']) - np.array(k['nose']))),\n",
    "                  np.sum(np.abs(np.array(k['mouth_left']) - np.array(k['mouth_right'])))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:33:41.290213Z",
     "start_time": "2019-12-26T15:33:41.286208Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "res = []\n",
    "for key in d.keys():\n",
    "    res.append(sum_distances(d[key]['keypoints']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:33:41.911213Z",
     "start_time": "2019-12-26T15:33:41.905213Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([111, 116,  69,  84]),\n",
       " array([122, 109,  71,  96]),\n",
       " array([116,  81, 101, 106]),\n",
       " array([95, 86, 95, 77]),\n",
       " array([115,  83, 104, 106]),\n",
       " array([104,  85, 101,  87]),\n",
       " array([ 90,  70, 110,  82]),\n",
       " array([103,  67, 102,  99]),\n",
       " array([107,  92,  89,  84]),\n",
       " array([100,  92,  78,  84]),\n",
       " array([87, 99, 76, 64]),\n",
       " array([ 99, 113,  80,  78])]"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:28:58.728260Z",
     "start_time": "2019-12-26T15:28:58.720273Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "k = obj['keypoints']\n",
    "first = np.array([np.sum(np.abs(np.array(k['left_eye']) - np.array(k['right_eye']))),\n",
    "                  np.sum(np.abs(np.array(k['left_eye']) - np.array(k['nose']))),\n",
    "                  np.sum(np.abs(np.array(k['right_eye']) - np.array(k['nose']))),\n",
    "                  np.sum(np.abs(np.array(k['mouth_left']) - np.array(k['mouth_right'])))])\n",
    "# first = np.array([k['left_eye'], k['right_eye'], k['nose'], k['mouth_left'], k['mouth_right']])\n",
    "k = obj2['keypoints']\n",
    "other = np.array([np.sum(np.abs(np.array(k['left_eye']) - np.array(k['right_eye']))),\n",
    "                  np.sum(np.abs(np.array(k['left_eye']) - np.array(k['nose']))),\n",
    "                  np.sum(np.abs(np.array(k['right_eye']) - np.array(k['nose']))),\n",
    "                  np.sum(np.abs(np.array(k['mouth_left']) - np.array(k['mouth_right'])))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:28:59.327751Z",
     "start_time": "2019-12-26T15:28:59.322782Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([111, 116,  69,  84])"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:29:00.047273Z",
     "start_time": "2019-12-26T15:29:00.042272Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([122, 109,  71,  96])"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "other"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:25:20.311820Z",
     "start_time": "2019-12-26T15:25:20.307800Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(588, 267)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "k['left_eye']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T15:25:26.598035Z",
     "start_time": "2019-12-26T15:25:26.593036Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(690, 287)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "k['right_eye']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T13:58:12.670744Z",
     "start_time": "2019-12-26T13:58:12.663749Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(77, 4)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['label'] == 'REAL'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T13:58:28.517440Z",
     "start_time": "2019-12-26T13:58:28.510442Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(323, 4)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['label'] == 'FAKE'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T14:01:30.231711Z",
     "start_time": "2019-12-26T14:01:30.195715Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "crops = load_obj('crops_fake')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T14:00:58.936053Z",
     "start_time": "2019-12-26T14:00:58.932051Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'../to_be_annotated_t/'"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'../to_be_annotated_{}/'.format('t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T14:01:32.766842Z",
     "start_time": "2019-12-26T14:01:31.829840Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'aagfhgtpmv.mp4': {'0_0': {'coordinates': {'x1': 338,\n",
       "    'y1': 167,\n",
       "    'width': 219,\n",
       "    'height': 291},\n",
       "   'keypoints': {'left_eye': (416, 279),\n",
       "    'right_eye': (514, 266),\n",
       "    'nose': (495, 316),\n",
       "    'mouth_left': (452, 393),\n",
       "    'mouth_right': (527, 384)},\n",
       "   'confidence': 0.9999223947525024},\n",
       "  '25_0': {'coordinates': {'x1': 503, 'y1': 173, 'width': 220, 'height': 286},\n",
       "   'keypoints': {'left_eye': (588, 267),\n",
       "    'right_eye': (690, 287),\n",
       "    'nose': (648, 316),\n",
       "    'mouth_left': (588, 382),\n",
       "    'mouth_right': (667, 399)},\n",
       "   'confidence': 0.9999939203262329},\n",
       "  '50_0': {'coordinates': {'x1': 458, 'y1': 177, 'width': 219, 'height': 290},\n",
       "   'keypoints': {'left_eye': (520, 276),\n",
       "    'right_eye': (624, 288),\n",
       "    'nose': (556, 321),\n",
       "    'mouth_left': (508, 389),\n",
       "    'mouth_right': (602, 401)},\n",
       "   'confidence': 0.9999035596847534},\n",
       "  '75_0': {'coordinates': {'x1': 851, 'y1': 202, 'width': 198, 'height': 267},\n",
       "   'keypoints': {'left_eye': (900, 306),\n",
       "    'right_eye': (992, 303),\n",
       "    'nose': (943, 349),\n",
       "    'mouth_left': (910, 412),\n",
       "    'mouth_right': (985, 410)},\n",
       "   'confidence': 0.9999959468841553},\n",
       "  '100_0': {'coordinates': {'x1': 1072,\n",
       "    'y1': 199,\n",
       "    'width': 204,\n",
       "    'height': 288},\n",
       "   'keypoints': {'left_eye': (1121, 300),\n",
       "    'right_eye': (1222, 314),\n",
       "    'nose': (1154, 350),\n",
       "    'mouth_left': (1106, 408),\n",
       "    'mouth_right': (1196, 424)},\n",
       "   'confidence': 0.9999788999557495},\n",
       "  '125_0': {'coordinates': {'x1': 1091,\n",
       "    'y1': 167,\n",
       "    'width': 206,\n",
       "    'height': 284},\n",
       "   'keypoints': {'left_eye': (1138, 278),\n",
       "    'right_eye': (1237, 273),\n",
       "    'nose': (1182, 319),\n",
       "    'mouth_left': (1147, 386),\n",
       "    'mouth_right': (1232, 384)},\n",
       "   'confidence': 0.9995755553245544},\n",
       "  '150_0': {'coordinates': {'x1': 1020,\n",
       "    'y1': 189,\n",
       "    'width': 192,\n",
       "    'height': 275},\n",
       "   'keypoints': {'left_eye': (1058, 292),\n",
       "    'right_eye': (1146, 294),\n",
       "    'nose': (1081, 339),\n",
       "    'mouth_left': (1056, 396),\n",
       "    'mouth_right': (1132, 402)},\n",
       "   'confidence': 0.9999614953994751},\n",
       "  '175_0': {'coordinates': {'x1': 1060,\n",
       "    'y1': 120,\n",
       "    'width': 206,\n",
       "    'height': 266},\n",
       "   'keypoints': {'left_eye': (1107, 222),\n",
       "    'right_eye': (1202, 230),\n",
       "    'nose': (1133, 263),\n",
       "    'mouth_left': (1092, 318),\n",
       "    'mouth_right': (1180, 329)},\n",
       "   'confidence': 0.9999535083770752},\n",
       "  '200_0': {'coordinates': {'x1': 925, 'y1': 110, 'width': 206, 'height': 269},\n",
       "   'keypoints': {'left_eye': (981, 220),\n",
       "    'right_eye': (1075, 207),\n",
       "    'nose': (1036, 257),\n",
       "    'mouth_left': (1005, 318),\n",
       "    'mouth_right': (1081, 310)},\n",
       "   'confidence': 0.999936580657959},\n",
       "  '225_0': {'coordinates': {'x1': 937, 'y1': 117, 'width': 199, 'height': 269},\n",
       "   'keypoints': {'left_eye': (992, 218),\n",
       "    'right_eye': (1085, 211),\n",
       "    'nose': (1049, 253),\n",
       "    'mouth_left': (1010, 318),\n",
       "    'mouth_right': (1090, 314)},\n",
       "   'confidence': 0.9998329877853394},\n",
       "  '250_0': {'coordinates': {'x1': 940, 'y1': 156, 'width': 183, 'height': 250},\n",
       "   'keypoints': {'left_eye': (1002, 251),\n",
       "    'right_eye': (1088, 250),\n",
       "    'nose': (1057, 295),\n",
       "    'mouth_left': (1018, 353),\n",
       "    'mouth_right': (1082, 353)},\n",
       "   'confidence': 0.9999988079071045},\n",
       "  '275_0': {'coordinates': {'x1': 618, 'y1': 142, 'width': 211, 'height': 281},\n",
       "   'keypoints': {'left_eye': (694, 252),\n",
       "    'right_eye': (790, 249),\n",
       "    'nose': (760, 299),\n",
       "    'mouth_left': (712, 359),\n",
       "    'mouth_right': (788, 357)},\n",
       "   'confidence': 0.9999850988388062}},\n",
       " 'aapnvogymq.mp4': {'0_0': {'coordinates': {'x1': 1189,\n",
       "    'y1': 192,\n",
       "    'width': 146,\n",
       "    'height': 185},\n",
       "   'keypoints': {'left_eye': (1209, 262),\n",
       "    'right_eye': (1248, 255),\n",
       "    'nose': (1210, 300),\n",
       "    'mouth_left': (1231, 338),\n",
       "    'mouth_right': (1260, 333)},\n",
       "   'confidence': 0.9999994039535522},\n",
       "  '0_1': {'coordinates': {'x1': 585, 'y1': 241, 'width': 126, 'height': 191},\n",
       "   'keypoints': {'left_eye': (664, 318),\n",
       "    'right_eye': (697, 316),\n",
       "    'nose': (707, 350),\n",
       "    'mouth_left': (673, 389),\n",
       "    'mouth_right': (696, 387)},\n",
       "   'confidence': 0.9999772310256958},\n",
       "  '25_0': {'coordinates': {'x1': 1183, 'y1': 184, 'width': 153, 'height': 190},\n",
       "   'keypoints': {'left_eye': (1217, 258),\n",
       "    'right_eye': (1269, 246),\n",
       "    'nose': (1233, 289),\n",
       "    'mouth_left': (1239, 333),\n",
       "    'mouth_right': (1279, 325)},\n",
       "   'confidence': 0.9999915361404419},\n",
       "  '25_1': {'coordinates': {'x1': 554, 'y1': 234, 'width': 137, 'height': 215},\n",
       "   'keypoints': {'left_eye': (641, 315),\n",
       "    'right_eye': (677, 307),\n",
       "    'nose': (694, 349),\n",
       "    'mouth_left': (656, 399),\n",
       "    'mouth_right': (685, 393)},\n",
       "   'confidence': 0.9998846054077148},\n",
       "  '50_0': {'coordinates': {'x1': 1181, 'y1': 190, 'width': 144, 'height': 186},\n",
       "   'keypoints': {'left_eye': (1196, 264),\n",
       "    'right_eye': (1241, 257),\n",
       "    'nose': (1199, 301),\n",
       "    'mouth_left': (1217, 342),\n",
       "    'mouth_right': (1253, 335)},\n",
       "   'confidence': 0.9999994039535522},\n",
       "  '50_1': {'coordinates': {'x1': 557, 'y1': 224, 'width': 142, 'height': 206},\n",
       "   'keypoints': {'left_eye': (629, 300),\n",
       "    'right_eye': (686, 304),\n",
       "    'nose': (679, 338),\n",
       "    'mouth_left': (634, 380),\n",
       "    'mouth_right': (683, 381)},\n",
       "   'confidence': 0.999972939491272},\n",
       "  '75_0': {'coordinates': {'x1': 1179, 'y1': 186, 'width': 143, 'height': 188},\n",
       "   'keypoints': {'left_eye': (1193, 263),\n",
       "    'right_eye': (1237, 256),\n",
       "    'nose': (1197, 299),\n",
       "    'mouth_left': (1214, 341),\n",
       "    'mouth_right': (1248, 335)},\n",
       "   'confidence': 0.9999995231628418},\n",
       "  '75_1': {'coordinates': {'x1': 551, 'y1': 225, 'width': 144, 'height': 200},\n",
       "   'keypoints': {'left_eye': (615, 305),\n",
       "    'right_eye': (675, 301),\n",
       "    'nose': (666, 336),\n",
       "    'mouth_left': (626, 378),\n",
       "    'mouth_right': (677, 373)},\n",
       "   'confidence': 0.9999712705612183},\n",
       "  '100_0': {'coordinates': {'x1': 1174,\n",
       "    'y1': 185,\n",
       "    'width': 146,\n",
       "    'height': 192},\n",
       "   'keypoints': {'left_eye': (1188, 261),\n",
       "    'right_eye': (1236, 254),\n",
       "    'nose': (1193, 298),\n",
       "    'mouth_left': (1209, 341),\n",
       "    'mouth_right': (1247, 335)},\n",
       "   'confidence': 0.9999994039535522},\n",
       "  '100_1': {'coordinates': {'x1': 550, 'y1': 212, 'width': 153, 'height': 224},\n",
       "   'keypoints': {'left_eye': (622, 302),\n",
       "    'right_eye': (682, 299),\n",
       "    'nose': (675, 334),\n",
       "    'mouth_left': (627, 380),\n",
       "    'mouth_right': (681, 374)},\n",
       "   'confidence': 0.9999619722366333},\n",
       "  '125_0': {'coordinates': {'x1': 1173,\n",
       "    'y1': 187,\n",
       "    'width': 143,\n",
       "    'height': 187},\n",
       "   'keypoints': {'left_eye': (1186, 261),\n",
       "    'right_eye': (1234, 253),\n",
       "    'nose': (1192, 297),\n",
       "    'mouth_left': (1207, 340),\n",
       "    'mouth_right': (1244, 333)},\n",
       "   'confidence': 0.9999996423721313},\n",
       "  '125_1': {'coordinates': {'x1': 535, 'y1': 238, 'width': 141, 'height': 198},\n",
       "   'keypoints': {'left_eye': (613, 312),\n",
       "    'right_eye': (666, 314),\n",
       "    'nose': (664, 351),\n",
       "    'mouth_left': (622, 392),\n",
       "    'mouth_right': (666, 392)},\n",
       "   'confidence': 0.9987745881080627},\n",
       "  '150_0': {'coordinates': {'x1': 1168,\n",
       "    'y1': 187,\n",
       "    'width': 143,\n",
       "    'height': 186},\n",
       "   'keypoints': {'left_eye': (1183, 263),\n",
       "    'right_eye': (1231, 253),\n",
       "    'nose': (1191, 298),\n",
       "    'mouth_left': (1205, 341),\n",
       "    'mouth_right': (1243, 333)},\n",
       "   'confidence': 0.9999995231628418},\n",
       "  '150_1': {'coordinates': {'x1': 534, 'y1': 236, 'width': 142, 'height': 206},\n",
       "   'keypoints': {'left_eye': (612, 313),\n",
       "    'right_eye': (667, 319),\n",
       "    'nose': (662, 353),\n",
       "    'mouth_left': (612, 395),\n",
       "    'mouth_right': (660, 397)},\n",
       "   'confidence': 0.9990805387496948},\n",
       "  '175_0': {'coordinates': {'x1': 1158,\n",
       "    'y1': 185,\n",
       "    'width': 145,\n",
       "    'height': 188},\n",
       "   'keypoints': {'left_eye': (1175, 262),\n",
       "    'right_eye': (1226, 253),\n",
       "    'nose': (1185, 299),\n",
       "    'mouth_left': (1197, 341),\n",
       "    'mouth_right': (1234, 333)},\n",
       "   'confidence': 0.9999995231628418},\n",
       "  '175_1': {'coordinates': {'x1': 563, 'y1': 219, 'width': 141, 'height': 237},\n",
       "   'keypoints': {'left_eye': (654, 308),\n",
       "    'right_eye': (686, 306),\n",
       "    'nose': (699, 350),\n",
       "    'mouth_left': (658, 397),\n",
       "    'mouth_right': (680, 396)},\n",
       "   'confidence': 0.9999889135360718},\n",
       "  '200_0': {'coordinates': {'x1': 1161,\n",
       "    'y1': 189,\n",
       "    'width': 158,\n",
       "    'height': 187},\n",
       "   'keypoints': {'left_eye': (1191, 274),\n",
       "    'right_eye': (1255, 262),\n",
       "    'nose': (1218, 310),\n",
       "    'mouth_left': (1212, 345),\n",
       "    'mouth_right': (1268, 333)},\n",
       "   'confidence': 0.9999996423721313},\n",
       "  '200_1': {'coordinates': {'x1': 545, 'y1': 225, 'width': 135, 'height': 223},\n",
       "   'keypoints': {'left_eye': (633, 306),\n",
       "    'right_eye': (672, 308),\n",
       "    'nose': (680, 345),\n",
       "    'mouth_left': (631, 388),\n",
       "    'mouth_right': (666, 390)},\n",
       "   'confidence': 0.9999912977218628},\n",
       "  '225_0': {'coordinates': {'x1': 1157,\n",
       "    'y1': 184,\n",
       "    'width': 141,\n",
       "    'height': 186},\n",
       "   'keypoints': {'left_eye': (1169, 256),\n",
       "    'right_eye': (1219, 251),\n",
       "    'nose': (1174, 292),\n",
       "    'mouth_left': (1187, 334),\n",
       "    'mouth_right': (1225, 331)},\n",
       "   'confidence': 0.9999996423721313},\n",
       "  '225_1': {'coordinates': {'x1': 526, 'y1': 220, 'width': 149, 'height': 211},\n",
       "   'keypoints': {'left_eye': (591, 301),\n",
       "    'right_eye': (655, 304),\n",
       "    'nose': (638, 336),\n",
       "    'mouth_left': (593, 375),\n",
       "    'mouth_right': (653, 376)},\n",
       "   'confidence': 0.9999985694885254},\n",
       "  '250_0': {'coordinates': {'x1': 1172,\n",
       "    'y1': 180,\n",
       "    'width': 142,\n",
       "    'height': 189},\n",
       "   'keypoints': {'left_eye': (1185, 254),\n",
       "    'right_eye': (1234, 248),\n",
       "    'nose': (1191, 292),\n",
       "    'mouth_left': (1202, 334),\n",
       "    'mouth_right': (1240, 330)},\n",
       "   'confidence': 0.9999998807907104},\n",
       "  '250_1': {'coordinates': {'x1': 535, 'y1': 227, 'width': 144, 'height': 200},\n",
       "   'keypoints': {'left_eye': (605, 305),\n",
       "    'right_eye': (663, 302),\n",
       "    'nose': (653, 343),\n",
       "    'mouth_left': (611, 383),\n",
       "    'mouth_right': (660, 381)},\n",
       "   'confidence': 0.9990209341049194},\n",
       "  '275_0': {'coordinates': {'x1': 1179,\n",
       "    'y1': 177,\n",
       "    'width': 146,\n",
       "    'height': 195},\n",
       "   'keypoints': {'left_eye': (1193, 252),\n",
       "    'right_eye': (1244, 248),\n",
       "    'nose': (1198, 292),\n",
       "    'mouth_left': (1210, 334),\n",
       "    'mouth_right': (1248, 332)},\n",
       "   'confidence': 0.9999992847442627},\n",
       "  '275_1': {'coordinates': {'x1': 535, 'y1': 226, 'width': 140, 'height': 213},\n",
       "   'keypoints': {'left_eye': (611"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<b>limit_output extension: Maximum message size of 10000 exceeded with 829018 characters</b>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "crops"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T12:34:16.991336Z",
     "start_time": "2019-12-26T12:34:16.973338Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>label</th>\n",
       "      <th>split</th>\n",
       "      <th>original</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>aczrgyricp.mp4</td>\n",
       "      <td>FAKE</td>\n",
       "      <td>train</td>\n",
       "      <td>slwkmefgde.mp4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            index label  split        original\n",
       "9  aczrgyricp.mp4  FAKE  train  slwkmefgde.mp4"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['index'] == 'aczrgyricp.mp4']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T01:59:14.761027Z",
     "start_time": "2019-12-26T01:59:14.753051Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# crops = load_obj('crops')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T00:24:44.840173Z",
     "start_time": "2019-12-26T00:24:44.837173Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "img = video.get(275)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T00:36:45.963464Z",
     "start_time": "2019-12-26T00:36:45.958493Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1080, 1920, 3)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T00:36:46.646645Z",
     "start_time": "2019-12-26T00:36:46.643645Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "h, w, d = img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T01:45:49.178047Z",
     "start_time": "2019-12-26T01:45:48.846053Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "detector = MTCNN()\n",
    "img = video.get(275)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T01:45:50.287048Z",
     "start_time": "2019-12-26T01:45:49.607047Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "results = detector.detect_faces(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T01:46:51.872196Z",
     "start_time": "2019-12-26T01:46:51.866188Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'box': [618, 142, 211, 281],\n",
       "  'confidence': 0.9999850988388062,\n",
       "  'keypoints': {'left_eye': (694, 252),\n",
       "   'right_eye': (790, 249),\n",
       "   'nose': (760, 299),\n",
       "   'mouth_left': (712, 359),\n",
       "   'mouth_right': (788, 357)}},\n",
       " {'box': [-5, 769, 120, 148],\n",
       "  'confidence': 0.7400879859924316,\n",
       "  'keypoints': {'left_eye': (26, 827),\n",
       "   'right_eye': (69, 820),\n",
       "   'nose': (39, 858),\n",
       "   'mouth_left': (33, 882),\n",
       "   'mouth_right': (69, 875)}}]"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-26T01:47:08.029191Z",
     "start_time": "2019-12-26T01:47:08.023194Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9999850988388062"
      ]
     },
     "execution_count": 227,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results[0]['confidence']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## Get faces from videos using CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:48:48.803034Z",
     "start_time": "2019-12-24T00:48:48.765012Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Load the cascade\n",
    "face_cascade = cv2.CascadeClassifier('C:\\\\Users\\\\Anti\\\\Anaconda3\\\\envs\\\\ml\\\\Library\\\\etc\\\\haarcascades\\\\haarcascade_frontalface_default.xml')\n",
    "\n",
    "def return_face_coordinates(img):\n",
    "    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n",
    "    # Detect faces\n",
    "    faces = face_cascade.detectMultiScale(gray, 1.1, 4)\n",
    "    \n",
    "    return faces\n",
    "\n",
    "def return_face_crops(img):\n",
    "    face_coordinates = return_face_coordinates(img)\n",
    "    crops = []\n",
    "    for (x, y, w, h) in face_coordinates:\n",
    "#         cv2.rectangle(img, (x, y), (x+w, y+h), (255, 0, 0), 2)\n",
    "        crops.append({'img': img[y:y+h, x:x+w], 'coordinates': {'x':x, 'y':y, 'w':w, 'h':h}})\n",
    "    \n",
    "    return crops\n",
    "\n",
    "def get_face_crops_from_video(video, frame_interval):\n",
    "    crops = {}\n",
    "    for frame in range(0, video.__len__(), frame_interval):\n",
    "        img = video.get(frame)\n",
    "        c = return_face_crops(img)\n",
    "        crops[frame] = c\n",
    "        \n",
    "    return crops\n",
    "\n",
    "def save_crops(path, video_file_name, crops):\n",
    "    os.makedirs(path, exist_ok=True)\n",
    "    for key in crops.keys():\n",
    "        cnt = 0\n",
    "        for crop in crops[key]:\n",
    "            imwrite(os.path.join(path, '{}_{}_{}.{}'.format(video_file_name, key, cnt, 'png')), crop['img'])\n",
    "            cnt = cnt + 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:49:01.028622Z",
     "start_time": "2019-12-24T00:48:48.804010Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aagfhgtpmv.mp4\n"
     ]
    }
   ],
   "source": [
    "crops = {}\n",
    "for fake_video_file_name in df[df['label'] == 'FAKE']['index'].to_numpy():\n",
    "    print(fake_video_file_name)\n",
    "    video = Video(os.path.join(videos_folder, fake_video_file_name))\n",
    "    crops = get_face_crops_from_video(video, 15)\n",
    "#     crops[fake_video_file_name] = get_face_crops_from_video(video, 30)\n",
    "    save_crops('../to_be_annotated/', fake_video_file_name.split('.')[0], crops)\n",
    "    break\n",
    "    # crops = get_face_crops_from_video(video, 30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:49:04.307676Z",
     "start_time": "2019-12-24T00:49:01.035629Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "save_crops('../to_be_annotated/', fake_video_file_name.split('.')[0], crops)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:49:04.313629Z",
     "start_time": "2019-12-24T00:49:04.309666Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "coordinates = {}\n",
    "for key in crops.keys():\n",
    "    coordinates[key] = []\n",
    "    for crop in crops[key]:\n",
    "        coordinates[key].append(crop['coordinates'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:49:04.331628Z",
     "start_time": "2019-12-24T00:49:04.315630Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: [{'x': 69, 'y': 1020, 'w': 58, 'h': 58},\n",
       "  {'x': 1703, 'y': 512, 'w': 108, 'h': 108},\n",
       "  {'x': 351, 'y': 173, 'w': 245, 'h': 245}],\n",
       " 15: [{'x': 1461, 'y': 511, 'w': 102, 'h': 102},\n",
       "  {'x': 415, 'y': 178, 'w': 272, 'h': 272},\n",
       "  {'x': 91, 'y': 738, 'w': 50, 'h': 50}],\n",
       " 30: [{'x': 1420, 'y': 663, 'w': 62, 'h': 62},\n",
       "  {'x': 1220, 'y': 505, 'w': 101, 'h': 101},\n",
       "  {'x': 1425, 'y': 977, 'w': 73, 'h': 73},\n",
       "  {'x': 502, 'y': 174, 'w': 271, 'h': 271},\n",
       "  {'x': 1689, 'y': 421, 'w': 77, 'h': 77}],\n",
       " 45: [{'x': 1312, 'y': 661, 'w': 60, 'h': 60},\n",
       "  {'x': 1321, 'y': 973, 'w': 65, 'h': 65},\n",
       "  {'x': 1111, 'y': 503, 'w': 100, 'h': 100},\n",
       "  {'x': 412, 'y': 177, 'w': 268, 'h': 268},\n",
       "  {'x': 1573, 'y': 417, 'w': 79, 'h': 79}],\n",
       " 60: [{'x': 1490, 'y': 503, 'w': 119, 'h': 119},\n",
       "  {'x': 617, 'y': 184, 'w': 268, 'h': 268}],\n",
       " 75: [{'x': 1114, 'y': 288, 'w': 58, 'h': 58},\n",
       "  {'x': 816, 'y': 202, 'w': 259, 'h': 259}],\n",
       " 90: [{'x': 942, 'y': 219, 'w': 274, 'h': 274},\n",
       "  {'x': 1366, 'y': 803, 'w': 186, 'h': 186},\n",
       "  {'x': 146, 'y': 910, 'w': 103, 'h': 103}],\n",
       " 105: [{'x': 767, 'y': 372, 'w': 34, 'h': 34},\n",
       "  {'x': 1059, 'y': 182, 'w': 287, 'h': 287},\n",
       "  {'x': 486, 'y': 903, 'w': 98, 'h': 98}],\n",
       " 120: [{'x': 50, 'y': 41, 'w': 91, 'h': 91},\n",
       "  {'x': 1068, 'y': 168, 'w': 285, 'h': 285},\n",
       "  {'x': 788, 'y': 898, 'w': 96, 'h': 96}],\n",
       " 135: [{'x': 378, 'y': 65, 'w': 70, 'h': 70},\n",
       "  {'x': 1016, 'y': 176, 'w': 256, 'h': 256}],\n",
       " 150: [{'x': 455, 'y': 10, 'w': 117, 'h': 117},\n",
       "  {'x': 1526, 'y': 454, 'w': 65, 'h': 65},\n",
       "  {'x': 830, 'y': 224, 'w': 110, 'h': 110},\n",
       "  {'x': 978, 'y': 197, 'w': 250, 'h': 250},\n",
       "  {'x': 899, 'y': 181, 'w': 54, 'h': 54},\n",
       "  {'x': 621, 'y': 72, 'w': 67, 'h': 67}],\n",
       " 165: [{'x': 1528, 'y': 920, 'w': 100, 'h': 100},\n",
       "  {'x': 1326, 'y': 579, 'w': 59, 'h': 59},\n",
       "  {'x': 674, 'y': 17, 'w': 117, 'h': 117},\n",
       "  {'x': 990, 'y': 152, 'w': 250, 'h': 250},\n",
       "  {'x': 1757, 'y': 461, 'w': 61, 'h': 61},\n",
       "  {'x': 835, 'y': 74, 'w': 73, 'h': 73},\n",
       "  {'x': 323, 'y': 446, 'w': 105, 'h': 105}],\n",
       " 180: [{'x': 1789, 'y': 937, 'w': 103, 'h': 103},\n",
       "  {'x': 1004, 'y': 113, 'w': 273, 'h': 273},\n",
       "  {'x': 223, 'y': 355, 'w': 187, 'h': 187}],\n",
       " 195: [{'x': 1396, 'y': 490, 'w': 69, 'h': 69},\n",
       "  {'x': 1067, 'y': 378, 'w': 88, 'h': 88},\n",
       "  {'x': 442, 'y': 367, 'w': 169, 'h': 169},\n",
       "  {'x': 951, 'y': 99, 'w': 256, 'h': 256},\n",
       "  {'x': 100, 'y': 455, 'w': 81, 'h': 81},\n",
       "  {'x': 1283, 'y': 78, 'w': 76, 'h': 76}],\n",
       " 210: [{'x': 536, 'y': 370, 'w': 169, 'h': 169},\n",
       "  {'x': 1619, 'y': 510, 'w': 28, 'h': 28},\n",
       "  {'x': 894, 'y': 110, 'w': 251, 'h': 251},\n",
       "  {'x': 1491, 'y': 491, 'w': 74, 'h': 74},\n",
       "  {'x': 203, 'y': 456, 'w': 81, 'h': 81},\n",
       "  {'x': 1479, 'y': 620, 'w': 60, 'h': 60},\n",
       "  {'x': 1381, 'y': 77, 'w': 77, 'h': 77}],\n",
       " 225: [{'x': 1600, 'y': 509, 'w': 29, 'h': 29},\n",
       "  {'x': 918, 'y': 116, 'w': 248, 'h': 248},\n",
       "  {'x': 1332, 'y': 599, 'w': 60, 'h': 60},\n",
       "  {'x': 99, 'y': 456, 'w': 66, 'h': 66},\n",
       "  {'x': 1474, 'y': 491, 'w': 76, 'h': 76},\n",
       "  {'x': 179, 'y': 453, 'w': 86, 'h': 86},\n",
       "  {'x': 510, 'y': 361, 'w': 178, 'h': 178},\n",
       "  {'x': 1459, 'y': 619, 'w': 65, 'h': 65},\n",
       "  {'x': 1362, 'y': 78, 'w': 77, 'h': 77}],\n",
       " 240: [{'x': 358, 'y': 920, 'w': 150, 'h': 150},\n",
       "  {'x': 302, 'y': 365, 'w': 172, 'h': 172},\n",
       "  {'x': 1011, 'y': 142, 'w': 243, 'h': 243}],\n",
       " 255: [{'x': 1754, 'y': 361, 'w': 27, 'h': 27},\n",
       "  {'x': 1446, 'y': 916, 'w': 99, 'h': 99},\n",
       "  {'x': 1336, 'y': 88, 'w': 119, 'h': 119},\n",
       "  {'x': 834, 'y': 161, 'w': 233, 'h': 233},\n",
       "  {'x': 1095, 'y': 972, 'w': 53, 'h': 53}],\n",
       " 270: [{'x': 661, 'y': 157, 'w': 255, 'h': 255},\n",
       "  {'x': 1141, 'y': 906, 'w': 94, 'h': 94}],\n",
       " 285: [{'x': 495, 'y': 129, 'w': 274, 'h': 274}]}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coordinates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:52:39.523806Z",
     "start_time": "2019-12-24T00:52:39.519766Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "from keras_vggface.vggface import VGGFace\n",
    "from keras_vggface.utils import preprocess_input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T01:01:25.560733Z",
     "start_time": "2019-12-24T01:01:25.557733Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "from mtcnn.mtcnn import MTCNN\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T01:01:26.327965Z",
     "start_time": "2019-12-24T01:01:26.322999Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# extract a single face from a given photograph\n",
    "def extract_face(img, required_size=(224, 224)):\n",
    "    # create the detector, using default weights\n",
    "    detector = MTCNN()\n",
    "    # detect faces in the image\n",
    "    results = detector.detect_faces(img)\n",
    "    # extract the bounding box from the first face\n",
    "    x1, y1, width, height = results[0]['box']\n",
    "    x2, y2 = x1 + width, y1 + height\n",
    "    # extract the face\n",
    "    face = img[y1:y2, x1:x2]\n",
    "    # resize pixels to the model size\n",
    "    image = Image.fromarray(face)\n",
    "    image = image.resize(required_size)\n",
    "    face_array = np.asarray(image)\n",
    "    return face_array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T01:02:15.279030Z",
     "start_time": "2019-12-24T01:02:14.248033Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "face = extract_face(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T01:03:58.182722Z",
     "start_time": "2019-12-24T01:03:58.006685Z"
    },
    "hidden": true,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x235a69da208>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(face)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:49:51.319932Z",
     "start_time": "2019-12-24T00:49:17.198312Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://github.com/rcmalli/keras-vggface/releases/download/v2.0/rcmalli_vggface_tf_senet50.h5\n",
      " 25239552/175688524 [===>..........................] - ETA: 31"
     ]
    },
    {
     "data": {
      "text/html": [
       "<b>limit_output extension: Maximum message size of 10000 exceeded with 10126 characters</b>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "vggface = VGGFace(model='senet50')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:51:34.533688Z",
     "start_time": "2019-12-24T00:51:34.528722Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Inputs: [<tf.Tensor 'input_1:0' shape=(None, 224, 224, 3) dtype=float32>]\n",
      "Outputs: [<tf.Tensor 'classifier/Softmax:0' shape=(None, 8631) dtype=float32>]\n"
     ]
    }
   ],
   "source": [
    "print('Inputs: %s' % vggface.inputs)\n",
    "print('Outputs: %s' % vggface.outputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:52:18.684050Z",
     "start_time": "2019-12-24T00:52:18.503055Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "img = video.get(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:53:24.331036Z",
     "start_time": "2019-12-24T00:53:24.325036Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 1080, 1920, 3)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.expand_dims(img, axis=0).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T01:02:30.814131Z",
     "start_time": "2019-12-24T01:02:30.809140Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "face_proccesed = preprocess_input(np.expand_dims(face.astype('float32'), axis=0), version=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T01:02:35.524334Z",
     "start_time": "2019-12-24T01:02:31.601835Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.9510621e-06, 5.7944007e-06, 9.5319929e-06, ..., 1.0868226e-06,\n",
       "        2.1479169e-08, 1.7774062e-06]], dtype=float32)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vggface.predict(face_proccesed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:27:39.175463Z",
     "start_time": "2019-12-24T00:27:39.170464Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "import face_recognition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:27:39.194462Z",
     "start_time": "2019-12-24T00:27:39.188465Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'aagfhgtpmv.mp4'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fake_video_file_name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:27:42.623462Z",
     "start_time": "2019-12-24T00:27:39.196465Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "batch_img = []\n",
    "for frame in range(0, video.__len__(), 15):\n",
    "    img = video.get(frame)\n",
    "    batch_img.append(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:27:42.896462Z",
     "start_time": "2019-12-24T00:27:42.624458Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1c6cfdac2e8>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-24T00:27:59.014703Z",
     "start_time": "2019-12-24T00:27:55.647632Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[(153, 608, 436, 325)],\n",
       " [(181, 658, 417, 423)],\n",
       " [(181, 730, 417, 494)],\n",
       " [(181, 658, 417, 423)],\n",
       " [(181, 850, 417, 614)],\n",
       " [(228, 1064, 464, 829)],\n",
       " [(228, 1184, 464, 948)],\n",
       " [(205, 1303, 440, 1067)],\n",
       " [(181, 1353, 464, 1070)],\n",
       " [(205, 1248, 402, 1051)],\n",
       " [(228, 1208, 464, 972)],\n",
       " [(157, 1232, 393, 996)],\n",
       " [(165, 1228, 362, 1031)],\n",
       " [(109, 1160, 345, 924)],\n",
       " [(125, 1108, 322, 912)],\n",
       " [(145, 1128, 342, 932)],\n",
       " [(165, 1208, 362, 1011)],\n",
       " [(165, 1029, 362, 832)],\n",
       " [(157, 873, 393, 638)],\n",
       " [(133, 730, 369, 494)]]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " face_recognition.batch_face_locations(batch_img, number_of_times_to_upsample=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T23:05:22.591025Z",
     "start_time": "2019-12-23T23:05:22.587025Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "20"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(batch_img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T23:05:18.544034Z",
     "start_time": "2019-12-23T23:05:17.260059Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "face_locations = face_recognition.face_locations(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T22:57:40.100924Z",
     "start_time": "2019-12-23T22:57:40.095923Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(217, 588, 440, 365)]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "face_locations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T22:03:57.031256Z",
     "start_time": "2019-12-23T22:03:57.023256Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'img': Array([[[253, 253, 253],\n",
       "          [253, 253, 253],\n",
       "          [255, 255, 255],\n",
       "          ...,\n",
       "          [204, 209, 187],\n",
       "          [204, 209, 187],\n",
       "          [204, 209, 187]],\n",
       "  \n",
       "         [[253, 253, 253],\n",
       "          [253, 253, 253],\n",
       "          [252, 252, 252],\n",
       "          ...,\n",
       "          [204, 209, 187],\n",
       "          [204, 209, 187],\n",
       "          [204, 209, 187]],\n",
       "  \n",
       "         [[253, 253, 253],\n",
       "          [253, 252, 255],\n",
       "          [252, 251, 254],\n",
       "          ...,\n",
       "          [204, 209, 189],\n",
       "          [204, 209, 189],\n",
       "          [204, 209, 187]],\n",
       "  \n",
       "         ...,\n",
       "  \n",
       "         [[239, 253, 164],\n",
       "          [238, 251, 167],\n",
       "          [236, 249, 165],\n",
       "          ...,\n",
       "          [204, 208, 191],\n",
       "          [204, 208, 191],\n",
       "          [204, 208, 191]],\n",
       "  \n",
       "         [[244, 253, 197],\n",
       "          [242, 251, 191],\n",
       "          [238, 247, 187],\n",
       "          ...,\n",
       "          [204, 208, 191],\n",
       "          [204, 208, 191],\n",
       "          [204, 208, 191]],\n",
       "  \n",
       "         [[250, 255, 203],\n",
       "          [248, 255, 197],\n",
       "          [246, 255, 195],\n",
       "          ...,\n",
       "          [204, 208, 191],\n",
       "          [204, 208, 191],\n",
       "          [204, 208, 191]]], dtype=uint8),\n",
       "  'coordinates': {'x': 1703, 'y': 512, 'w': 108, 'h': 108}},\n",
       " {'img': Array([[[126, 107,  95],\n",
       "          [118, 106,  96],\n",
       "          [112, 100,  90],\n",
       "          ...,\n",
       "          [112,  91,  80],\n",
       "          [115,  94,  83],\n",
       "          [117,  95,  87]],\n",
       "  \n",
       "         [[130, 111,  99],\n",
       "          [121, 109,  99],\n",
       "          [114, 102,  92],\n",
       "          ...,\n",
       "          [109,  88,  77],\n",
       "          [111,  90,  79],\n",
       "          [113,  91,  83]],\n",
       "  \n",
       "         [[132, 113, 101],\n",
       "          [123, 111, 101],\n",
       "          [117, 105,  95],\n",
       "          ...,\n",
       "          [100,  83,  73],\n",
       "          [103,  86,  76],\n",
       "          [107,  90,  82]],\n",
       "  \n",
       "         ...,\n",
       "  \n",
       "         [[ 91,  75,  60],\n",
       "          [ 97,  76,  63],\n",
       "          [102,  81,  68],\n",
       "          ...,\n",
       "          [ 94,  77,  67],\n",
       "          [ 92,  75,  65],\n",
       "          [ 92,  75,  67]],\n",
       "  \n",
       "         [[ 92,  76,  61],\n",
       "          [ 98,  77,  64],\n",
       "          [103,  82,  69],\n",
       "          ...,\n",
       "          [ 88,  71,  61],\n",
       "          [ 85,  68,  58],\n",
       "          [ 85,  68,  60]],\n",
       "  \n",
       "         [[ 94,  78,  63],\n",
       "          [100,  79,  66],\n",
       "          [104,  83,  70],\n",
       "          ...,\n",
       "          [ 85,  68,  58],\n",
       "          [ 80,  63,  53],\n",
       "          [ 80,  63,  55]]], dtype=uint8),\n",
       "  'coordinates': {'x': 69, 'y': 1020, 'w': 58, 'h': 58}},\n",
       " {'img': Array([[[119, 110,  93],\n",
       "          [116, 107,  92],\n",
       "          [114, 105,  90],\n",
       "          ...,\n",
       "          [ 85, 117,  86],\n",
       "          [ 84, 116,  87],\n",
       "          [ 82, 114,  85]],\n",
       "  \n",
       "         [[118, 109,  92],\n",
       "          [114, 105,  90],\n",
       "          [109, 100,  85],\n",
       "          ...,\n",
       "          [ 92, 124,  93],\n",
       "          [ 87, 119,  90],\n",
       "          [ 83, 115,  86]],\n",
       "  \n",
       "         [[110, 101,  84],\n",
       "          [108,  99,  84],\n",
       "          [104,  95,  80],\n",
       "          ...,\n",
       "          [ 92, 124,  93],\n",
       "          [ 87, 119,  90],\n",
       "          [ 83, 115,  86]],\n",
       "  \n",
       "         ...,\n",
       "  \n",
       "         [[ 66,  55,  50],\n",
       "          [ 76,  59,  49],\n",
       "          [ 80,  63,  53],\n",
       "          ...,\n",
       "          [218, 199, 142],\n",
       "          [216, 197, 140],\n",
       "          [216, 197, 140]],\n",
       "  \n",
       "         [[ 70,  54,  48],\n",
       "          [ 76,  59,  51],\n",
       "          [ 80,  63,  55],\n",
       "          ...,\n",
       "          [217, 201, 143],\n",
       "          [214, 198, 140],\n",
       "          [214, 198, 140]],\n",
       "  \n",
       "         [[ 71,  55,  49],\n",
       "          [ 76,  59,  51],\n",
       "          [ 79,  62,  54],\n",
       "          ...,\n",
       "          [218, 202, 144],\n",
       "          [216, 200, 142],\n",
       "          [214, 198, 140]]], dtype=uint8),\n",
       "  'coordinates': {'x': 351, 'y': 173, 'w': 245, 'h': 245}}]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crops[key]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:54:25.577201Z",
     "start_time": "2019-12-23T21:54:25.573201Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'aagfhgtpmv'"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fake_video_file_name.split('.')[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:46:50.004600Z",
     "start_time": "2019-12-23T21:46:49.996601Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys([0, 30, 60, 90, 120, 150, 180, 210, 240, 270])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crops['aagfhgtpmv.mp4'].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "scipy.misc.imsave(os.path.join('../to_be_annotated/'), image_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:39:44.098629Z",
     "start_time": "2019-12-23T21:39:43.809604Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "video = Video(os.path.join(videos_folder, videos[5]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:39:48.746122Z",
     "start_time": "2019-12-23T21:39:44.218605Z"
    },
    "hidden": true
   },
   "outputs": [],
   "source": [
    "crops = get_face_crops_from_video(video, 30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:39:48.874045Z",
     "start_time": "2019-12-23T21:39:48.870044Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(crops)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:43:12.089990Z",
     "start_time": "2019-12-23T21:43:12.084987Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(crops[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:39:49.007046Z",
     "start_time": "2019-12-23T21:39:49.002048Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "150"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(crops[0][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:39:49.134048Z",
     "start_time": "2019-12-23T21:39:49.130045Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(150, 150, 3)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crops[0][0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:43:47.813745Z",
     "start_time": "2019-12-23T21:43:47.684743Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1cf1554da20>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(crops[0][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-12-23T21:43:50.167226Z",
     "start_time": "2019-12-23T21:43:50.033226Z"
    },
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x1cf14d01198>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(crops[0][1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "position": {
    "height": "251px",
    "left": "992px",
    "right": "20px",
    "top": "127px",
    "width": "338px"
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}