Skip to content

Latest commit

 

History

History
executable file
·
25 lines (20 loc) · 759 Bytes

README.md

File metadata and controls

executable file
·
25 lines (20 loc) · 759 Bytes

EP-Net:More efficient pose estimation network with the classification-based keypoints detection

This repository contains a pytorch implementation of "EP-Net:More efficient pose estimation network with the classification-based keypoints detection".

This codebase provides:

  • test code
  • train code
  • dataset process code

installation

Check INSTALL.md for installation instructions.

train

reconfig the config file in E2P/configs/caffe2/keypoints_R_101_FPN.yaml

python E2P/tools/train_net.py

test

test code in E2P/pose_test

visualization

  • LINEMOD dataset: 'linemod'
  • Occ-LINEMOD dataset: 'occ-linemod'
  • YCB-Video: 'YCB-Video'