mkocabas/EpipolarPose

Self-Supervised Learning of 3D Human Pose using Multi-view Geometry (CVPR2019)

48
/ 100
Emerging

This project helps researchers and practitioners in fields like motion analysis or animation by accurately estimating 3D human poses from standard RGB images or video. It takes a single image of a person as input and outputs a precise 3D model of their pose, without needing special 3D sensing equipment or complex camera setups for training. This tool is ideal for scientists studying human movement or developers creating realistic character animations.

609 stars. No commits in the last 6 months.

Use this if you need to determine a person's 3D body posture from standard 2D video footage or still images, especially when 3D ground-truth data or camera calibration is unavailable.

Not ideal if your primary need is 2D pose detection, or if you require real-time performance on resource-constrained devices without dedicated GPUs.

human-pose-estimation motion-capture computer-vision biomechanics animation
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 22 / 25

How are scores calculated?

Stars

609

Forks

95

Language

Jupyter Notebook

License

Last pushed

Jun 26, 2019

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/mkocabas/EpipolarPose"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.