Skip to content

Real-time human pose estimation code based on pose proposal network(ECCV2018)

Notifications You must be signed in to change notification settings

noirmist/Pytorch_Pose_Proposal_Network

Repository files navigation

Pytorch_Pose_Proposal_Network

Real-time human pose estimation code based on pose proposal network(ECCV2018)

Webcam Test Result

Test result

Difference from Pose Proposal Network(T. Sekii- ECCV 2018)

Network - Dilated Residual Network

Network image

Training - GradNorm

GradNorm

Environment

OS: Ubuntu 16.04
Python version: 3.5.2
PyTorch version: 1.1.0

How To Use

  1. Data preprocessing by using code in data folder
  2. Update directory paths in train.sh
  3. Train the network
  4. Update directory paths in rt_test.sh and run the code

Reference

  • Pose proposal network: here
  • Evaluation code: here

About

Real-time human pose estimation code based on pose proposal network(ECCV2018)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published