MediaPipe-Face-Detection

Detect faces and locate facial features in real-time video and image streams.

Designed for sub-millisecond processing, this model predicts bounding boxes and pose skeletons (left eye, right eye, nose tip, mouth, left eye tragion, and right eye tragion) of faces in an image.