3D Human Pose Estimation in RGBD Images for Robotic Task Learning
This work addresses the problem of enabling service robots to learn tasks from human demonstrations without markers, which is incremental as it builds on existing keypoint detectors and depth integration.
The paper tackles 3D human pose estimation from single RGBD images, achieving performance that exceeds monocular color-based and depth-only methods, and integrates it with a learning from demonstration framework to enable a PR2 robot to imitate human manipulation actions in real-world settings.
We propose an approach to estimate 3D human pose in real world units from a single RGBD image and show that it exceeds performance of monocular 3D pose estimation approaches from color as well as pose estimation exclusively from depth. Our approach builds on robust human keypoint detectors for color images and incorporates depth for lifting into 3D. We combine the system with our learning from demonstration framework to instruct a service robot without the need of markers. Experiments in real world settings demonstrate that our approach enables a PR2 robot to imitate manipulation actions observed from a human teacher.