Human Motion Imitation for Humanoid by Recurrent Neural Network
This article is accepted in The 13th International Conference on Ubiquitous Robots and Ambient Intelligene (URAI).
Tags: publications Imitation
This article is accepted in The 13th International Conference on Ubiquitous Robots and Ambient Intelligene (URAI).
Tags: publications Imitation
In this paper, it is investigated how a humanoid can be controlled from a human motion in a tele-operation system. The proposed method is based on Recurrent Neural Networks (RNN) to extract features in nonlinear sequential data. Therefore, our proposed RNN model can extract the features of the relation between a human and a robot, and generate motion in a robot using these features. Also, by regarding the balancing state of the robot, the robot’s joints corresponding the operator is modified for stable contact against the ground. The effectiveness of the proposed new mapping algorithm is verified through simulations.
Please Visit Imitation project.