leoxiaobin/deep-high-resolution-net.pytorch
The project is an official implementation of our CVPR2019 paper "Deep High-Resolution Representation Learning for Human Pose Estimation"
Maintains high-resolution representations throughout the network via parallel multi-resolution subnetworks with repeated cross-scale fusion, rather than recovering them from low-resolution features. Built in PyTorch and optimized for NVIDIA GPUs, it achieves superior keypoint localization accuracy on COCO and MPII benchmarks while maintaining lower computational cost than ResNet-based baselines. Supports both top-down pose estimation and integrates with external person detectors for end-to-end pipeline deployment.
4,466 stars. No commits in the last 6 months.
Stars
4,466
Forks
926
Language
Cuda
License
MIT
Category
Last pushed
Aug 30, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/leoxiaobin/deep-high-resolution-net.pytorch"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
talmolab/sleap
A deep learning framework for multi-animal pose tracking.
open-mmlab/mmaction2
OpenMMLab's Next Generation Video Understanding Toolbox and Benchmark
jgraving/DeepPoseKit
a toolkit for pose estimation using deep learning
kennymckormick/pyskl
A toolbox for skeleton-based action recognition.
CMU-Perceptual-Computing-Lab/openpose_train
Training repository for OpenPose