Skip to content

Droliven/pymaf_reimplementation

Repository files navigation

Pymaf reimplementation

Levon 2021.12.27

reimplementation of Pyramidal Mesh Alignment Feedback (PyMAF) (ICCV 2021 Oral) for the 3D human mesh reconstruction task, projrct, paper, code.

Overview of our approach Pyramidal Mesh Alignment Feedback (PyMAF)

This reimplementation project has been linked to the corresponding paperwithcode page and the original arxiv page.

bugs to fix

由于 licence 问题导致的 h36m mosh 数据无法获得

h36m/
    h36m.npy
    h36m_mosh_train.npz
    h36m_mosh_valid_p2.npz
mkdir data
cd data
wget http://visiondata.cis.upenn.edu/volumetric/h36m/h36m_annot.tar
tar -xf h36m_annot.tar
rm h36m_annot.tar

# Download H36M images
mkdir -p h36m/images
cd h36m/images
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S1.tar
tar -xf S1.tar
rm S1.tar
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S5.tar
tar -xf S5.tar
rm S5.tar
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S6.tar
tar -xf S6.tar
rm S6.tar
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S7.tar
tar -xf S7.tar
rm S7.tar
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S8.tar
tar -xf S8.tar
rm S8.tar
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S9.tar
tar -xf S9.tar
rm S9.tar
wget http://visiondata.cis.upenn.edu/volumetric/h36m/S11.tar
tar -xf S11.tar
rm S11.tar
cd ../../..

异常复杂的数据情况

train

  • center, scale 共同决定 bounding box,
  • pose, shape 是 SMPL 参数
  • part 是 2D keypoint, 有 24 个点, 但也是三维的 ; 相应 openpose 也是 2D keypoint 但有 25 个点
  • S 是 3D skeleton, 有 24 个点, 但是四维的
DS length imgname (str) center (2) scale (1) pose (72) shape (10) part (24, 3) 2D 关节点 S (24, 4) 3D关节点 openpose has_smpl maskname partname gender
h36m 312188 S1/S1_Directions_1.54138969/S1_Directions_1.54138969_000001.jpg 1 1 1 1 1 1
mpii 14810 (unique 9899) images/015601864.jpg 1 1 1 1
lsp-orig 1000 images/im0001.jpg 1 1 1 1
lspet 9428 im00001.png 1 1 1 1
mpi-inf-3dhp 96507 S1/Seq1/imageFrames/video_0/frame_000001.jpg 1 1 1 1 1 1 1 1
coco 28344(unique 18127) train2014/COCO_train2014_000000044474.jpg 1 1 1 1

eval

DS length imgname center scale pose shape part S openpose has_smpl maskname partname gender
h36m-p1 109867 S9/S9_Directions_1.54138969/S9_Directions_1.54138969_000001.jpg 1 1 1
h36m-p2 27558 S9/S9_Directions_1.60457274/S9_Directions_1.60457274_000001.jpg 1 1 1
h36m-p2-mosh 26859 S9/S9_Directions_1.60457274/S9_Directions_1.60457274_000001.jpg 1 1 1 1 1 1
lsp 1000 images/im1001.jpg 1 1 1 1 1
mpi-inf-3dhp 2929 TS1/imageSequence/img_000001.jpg 1 1 1 1
coco 50197(unique 19010) val2014/COCO_val2014_000000537548.jpg 1 1 1
3dpw 35515(unique 24547) imageFiles/downtown_enterShop_00/image_00000.jpg 1 1 1 1 1

Bugs to fix

  • 当设置 mixture 时出现问题,使得 LSP_ori 等不包含 pose, shape, has_smpl 的数据集,出现了 has_sml datas 中存在问题, 问题出在了 runner.FitsDict
  • OpenDRRander 替换为了 pyRender
  • 用 20000 数据训练之后发现,eval 效果太差,而且人偏瘦。
  • 原版模型测试可以达到论文标称结果,因为之前 test() 里面没有写 self.model.eval().

About

reimplementation of PyMAF (ICCV 2021 Oral)

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages