Deep Learning with Predictive Control for Human Motion Tracking

2018/08/07 by Don Joven Agravante, Giovanni De Magistris, Agravante, Don Joven +7
Computer Science · #Human Pose and Action Recognition #Multimodal Machine Learning Applications #Advanced Vision and Imaging

paper · pdf · doi:10.48550/arxiv.1808.02200

Abstract

We propose to combine model predictive control with deep learning for the task of accurate human motion tracking with a robot. We design the MPC to allow switching between the learned and a conservative prediction. We also explored online learning with a DyBM model. We applied this method to human handwriting motion tracking with a UR-5 robot. The results show that the framework significantly improves tracking performance.

Related

• Deep Predictive Learning: Motion Learning Concept inspired by Cognitive Robotics
• DR-MPC: Deep Residual Model Predictive Control for Real-world Social Navigation
• Deep Sensorimotor Control by Imitating Predictive Models of Human Motion
• Achieving Faster and More Accurate Operation of Deep Predictive Learning
• Deep Visual Foresight for Planning Robot Motion