BlazePose: On-device Real-time Body Pose tracking
This enables real-time applications like fitness tracking and sign language recognition on mobile devices, representing an incremental improvement in efficiency.
The paper tackles real-time human pose estimation on mobile devices, achieving over 30 frames per second on a Pixel 2 phone with 33 body keypoints.
We present BlazePose, a lightweight convolutional neural network architecture for human pose estimation that is tailored for real-time inference on mobile devices. During inference, the network produces 33 body keypoints for a single person and runs at over 30 frames per second on a Pixel 2 phone. This makes it particularly suited to real-time use cases like fitness tracking and sign language recognition. Our main contributions include a novel body pose tracking solution and a lightweight body pose estimation neural network that uses both heatmaps and regression to keypoint coordinates.