ZheC/Realtime_Multi-Person_Pose_Estimation
A bottom-up deep learning approach for real-time multi-person human pose estimation published at CVPR 2017.
Velocity · 7d
+1.5
★ / day
Trend
→steady
star history
This repository contains the reference implementation of the CVPR 2017 oral paper on multi-person pose estimation. The approach uses a deep neural network to detect human body keypoints in images without requiring a separate person detector, operating in real-time. The code supports Caffe, Python, MATLAB, and C++11 implementations for training and inference.