OpenPose

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

34493 stars · 8040 forks observed · 2026-10-10

What it is AI-extracted, prompt v1, taxonomy v1, 2026-09-28, confidence 0.98

OpenPose is an open‑source C++ library that provides real‑time multi‑person keypoint detection for body, face, hands, and feet. It leverages deep‑learning models and runs on Linux, macOS, and Windows.

Use cases

• extract 2D body keypoints from live video streams
• track hand and facial landmarks for gesture recognition
• analyze full‑body motion in sports or dance recordings
• generate pose data for animation or VR avatars
• perform multi‑person pose estimation in surveillance footage
• feed keypoint coordinates into downstream AI pipelines

When to choose

• You need real‑time, high‑accuracy pose estimation for multiple people simultaneously.
• Your project can accommodate a C++ dependency stack (Caffe, OpenCV).
• You are working on research or applications that require detailed body, hand, face, and foot keypoints.

When to avoid

• You require a lightweight, mobile‑oriented solution with minimal runtime overhead.
• Your stack is limited to pure Python or JavaScript without native extensions.
• You need out‑of‑the‑box support for 3D pose or depth sensors without additional integration.

Facets

library · maturity active

computer-vision deep-learning image-processing artificial-intelligence computer-vision deep-learning windows real-time multi-person keypoint-detection body-pose hand-pose face-pose foot-pose c++ caffe opencv linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="CMU-Perceptual-Computing-Lab/openpose")