human pose estimation
OpenPose is a real-time multi-person 2D keypoint detection library that estimates body, face, hands, and feet keypoints (135 total) from single images using bottom-up Part Affinity Fields.
Solves the same problem with a different architecture or at a different layer. Expect to rewrite the integration.
Top-down multi-person pose and tracking with better accuracy on standard benchmarks, if you prefer top-down over OpenPose's speed.
Real-Time and Accurate Full-Body Multi-Person Pose Estimation&Tracking System
High-level Python pose library with pluggable models and clean API, choose for flexibility over OpenPose's monolithic C++.
Library for Fast and Flexible Human Pose Estimation
Official top-down Simple Baseline in PyTorch, choose for easy-to-modify research code but unmaintained and less integrated.
The project is an official implement of our ECCV2018 paper "Simple Baselines for Human Pose Estimation and Tracking(https://arxiv.org/abs/1804.06208)"
HRNet top-down pose model with strong accuracy, choose if you need high precision and can run a person detector first.
The project is an official implementation of our CVPR2019 paper "Deep High-Resolution Representation Learning for Human Pose Estimation"
Whole-body keypoint baseline with official COCO-WholeBody evaluation, choose for benchmark-driven whole-body work but top-down and research-oriented.
ECCV2020 paper "Whole-Body Human Pose Estimation in the Wild"
Bottom-up scale-aware pose estimator with better accuracy than OpenPose, choose if you can accept slower inference and no maintained releases.
This is an official implementation of our CVPR 2020 paper "HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation" (https://arxiv.org/abs/1908.10357)
Modern whole-body pose estimator with distilled RTMPose, choose for accurate top-down whole-body keypoints but requires PyTorch/MMPose stack.
"Effective Whole-body Pose Estimation with Two-stages Distillation" (ICCV 2023, CV4Metaverse Workshop)
Transformer-based top-down pose estimation with state-of-the-art accuracy, choose for high-precision experiments with GPU resources.
The official repo for [NeurIPS'22] "ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation" and [TPAMI'23] "ViTPose++: Vision Transformer for Generic Body Pose Estimation"
An earlier or more limited way to do the job. Still the right call when you need something small, proven, or CPU-only.
Original CVPR17 Caffe implementation of the PAF paper, unmaintained, for reproducing the baseline.
Code repo for realtime multi-person pose estimation in CVPR'17 (Oral)
Lightweight PyTorch model for CPU real-time pose, choose if you must run on CPU and can accept lower accuracy and fewer features.
Fast and accurate human pose estimation in PyTorch. Contains implementation of "Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose" paper.
TensorFlow implementation of Convolutional Pose Machines, an older single-person approach for CPM research.
Cascaded Pyramid Network in TensorFlow for top-down pose, choose for a classic CPN baseline but outdated and unmaintained.
[CVPR 2018] Cascaded Pyramid Network for Multi-Person Pose Estimation
These projects were analysed and named openpose among their alternatives. The relationship is not symmetric — how openpose rates them is a separate judgement, made when openpose is analysed in its own right.
calls openpose “The original bottom-up real-time keypoint library; prefer it for C++ integration and body+hand+face detection with a different architecture.”
Real-Time and Accurate Full-Body Multi-Person Pose Estimation&Tracking System
calls openpose “Choose when your need is specifically real-time multi-person 2D body/face/hand pose; bottom-up PAF output is more pose-focused and does not give object boxes/3D.”
Object detection, 3D detection, and pose estimation using center point detection:
calls openpose “choose if you need production-grade multi-person 2D keypoints with body/hand/face support, but it only covers pose.”
Ultralytics YOLO26, YOLO11, YOLOv8 — object detection, instance segmentation, semantic segmentation, image classification, pose estimation, object tracking