Hand-Object Interaction Controller (HOIC): Deep Reinforcement Learning for Reconstructing Interactions with Physics
This work addresses the challenge of faithfully simulating complex hand-object interactions for applications in robotics and virtual reality, representing an incremental advance in physics-based motion reconstruction.
The paper tackles the problem of reconstructing hand-object interaction motions from a single RGBD camera using deep reinforcement learning to incorporate physics, achieving improved reconstruction accuracy and physical correctness without heuristic rules.
Hand manipulating objects is an important interaction motion in our daily activities. We faithfully reconstruct this motion with a single RGBD camera by a novel deep reinforcement learning method to leverage physics. Firstly, we propose object compensation control which establishes direct object control to make the network training more stable. Meanwhile, by leveraging the compensation force and torque, we seamlessly upgrade the simple point contact model to a more physical-plausible surface contact model, further improving the reconstruction accuracy and physical correctness. Experiments indicate that without involving any heuristic physical rules, this work still successfully involves physics in the reconstruction of hand-object interactions which are complex motions hard to imitate with deep reinforcement learning. Our code and data are available at https://github.com/hu-hy17/HOIC.