5.4 池化层.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.0.0\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "print(tf.__version__)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pool2d(X, pool_size, mode='max'):\n",
    "    p_h, p_w = pool_size\n",
    "    Y = tf.zeros((X.shape[0] - p_h + 1, X.shape[1] - p_w +1))\n",
    "    Y = tf.Variable(Y)\n",
    "    for i in range(Y.shape[0]):\n",
    "        for j in range(Y.shape[1]):\n",
    "            if mode == 'max':\n",
    "                Y[i,j].assign(tf.reduce_max(X[i:i+p_h, j:j+p_w]))\n",
    "            elif mode =='avg':\n",
    "                Y[i,j].assign(tf.reduce_mean(X[i:i+p_h, j:j+p_w]))\n",
    "    return Y\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Variable 'Variable:0' shape=(2, 2) dtype=float32, numpy=\n",
       "array([[4., 5.],\n",
       "       [7., 8.]], dtype=float32)>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = tf.constant([[0,1,2],[3,4,5],[6,7,8]],dtype=tf.float32)\n",
    "pool2d(X, (2,2))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Variable 'Variable:0' shape=(2, 2) dtype=float32, numpy=\n",
       "array([[2., 3.],\n",
       "       [5., 6.]], dtype=float32)>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pool2d(X, (2,2), 'avg')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: id=112, shape=(1, 4, 4, 1), dtype=int32, numpy=\n",
       "array([[[[ 0],\n",
       "         [ 1],\n",
       "         [ 2],\n",
       "         [ 3]],\n",
       "\n",
       "        [[ 4],\n",
       "         [ 5],\n",
       "         [ 6],\n",
       "         [ 7]],\n",
       "\n",
       "        [[ 8],\n",
       "         [ 9],\n",
       "         [10],\n",
       "         [11]],\n",
       "\n",
       "        [[12],\n",
       "         [13],\n",
       "         [14],\n",
       "         [15]]]])>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# padding and stride\n",
    "#tensorflow default data_format == 'channels_last'\n",
    "#so (1,4,4,1) instead of (1,1,4,4)\n",
    "X = tf.reshape(tf.constant(range(16)), (1,4,4,1))\n",
    "X\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: id=113, shape=(1, 1, 1, 1), dtype=int32, numpy=array([[[[10]]]])>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pool2d = tf.keras.layers.MaxPool2D(pool_size=[3,3])\n",
    "pool2d(X)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: id=114, shape=(1, 2, 2, 1), dtype=int32, numpy=\n",
       "array([[[[10],\n",
       "         [11]],\n",
       "\n",
       "        [[14],\n",
       "         [15]]]])>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#I guess no custom padding settings in keras.layers?\n",
    "pool2d = tf.keras.layers.MaxPool2D(pool_size=[3,3],padding='same',strides=2)\n",
    "pool2d(X)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: id=120, shape=(2, 2, 2, 1), dtype=int32, numpy=\n",
       "array([[[[ 5],\n",
       "         [ 6]],\n",
       "\n",
       "        [[ 7],\n",
       "         [ 8]]],\n",
       "\n",
       "\n",
       "       [[[13],\n",
       "         [14]],\n",
       "\n",
       "        [[15],\n",
       "         [16]]]])>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 多通道\n",
    "X = tf.stack([X, X+1], axis=3)\n",
    "X = tf.reshape(X, (2,4,4,1))\n",
    "X.shape\n",
    "\n",
    "pool2d = tf.keras.layers.MaxPool2D(3, padding='same', strides=2)\n",
    "pool2d(X)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (tf2)",
   "language": "python",
   "name": "tf2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}