BlazePose: On-device Real-time Body Pose tracking

Abstract

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.

4 pages, 6 figures; CVPR Workshop on Computer Vision for Augmented and Virtual Reality, Seattle, WA, USA, 2020

References in corpus (2)

• BlazeFace: Sub-millisecond Neural Face Detection on Mobile GPUs
• Real-time Facial Surface Geometry from Monocular Video on Mobile GPUs

Cited by in corpus (5)

• Deep Learning based Virtual Point Tracking for Real-Time Target-less Dynamic Displacement Measurement in Railway Applications
• Unsupervised Motion Retargeting for Human-Robot Imitation
• Analyzing Data Efficiency and Performance of Machine Learning Algorithms for Assessing Low Back Pain Physical Rehabilitation Exercises
• Deep-Learning-Based Markerless Pose Estimation Systems in Gait Analysis: DeepLabCut Custom Training and the Refinement Function
• On-device Real-time Hand Gesture Recognition