CMU-Perceptual-Computing-Lab/openpose_train
Training code for OpenPose, a real-time multi-person human pose estimation model that detects 135 body keypoints.

Velocity · 7d
+0.2
★ / day
Trend
→steady
star history
This repository contains the experimental training code for the OpenPose system, which is a real-time multi-person pose estimation model based on deep learning. The system jointly detects human body, hand, facial, and foot keypoints on images. This training repository complements the main OpenPose inference repository and is used to train the pose estimation models.