OpenPose multi-person face, body and hand keypoint detection demo

Perceptual Computing Laboratory has released a short video demonstrating the detection of faces, bodies and hands for multiple people using OpenPose, a C++ library created using OpenCV and Caffe. According to its authors, OpenPose “[…] represents the first real-time system to jointly detect human body, hand and facial keypoints (in total 130 keypoints) on single images”.  Main features of the library include:

• Multi-person 15 or 18-keypoint body pose estimation and rendering. Running time invariant to number of people on the image.
• Multi-person 2×21-keypoint hand estimation and rendering.
• Multi-person 70-keypoint face estimation and rendering.
• Flexible and easy to configure multi-threading module.
• Image, video, and webcam reader.
• Ability to save and load the results in various formats (JSON, XML, PNG, JPG, …).
• Small display and GUI for simple result visualisation.
• All the functionality is wrapped into a simple-to-use OpenPose Wrapper class.

OpenPose if free for non-commercial use. See a demo on YouTube and download the source code and find out more on GitHub.
 

Face tracking

 

Multi-person keypoint detection