Volume 10, Number 5

An Application of Convolutional Neural Networks on Human Intention Prediction


Lin Zhang1, Shengchao Li2, Hao Xiong2, Xiumin Diao2 and Ou Ma1, 1University of Cincinnati, USA and 2Purdue University, USA


Due to the rapidly increasing need of human-robot interaction (HRI), more intelligent robots are in demand. However, the vast majority of robots can only follow strict instructions, which seriously restricts their flexibility and versatility. A critical fact that strongly negates the experience of HRI is that robots cannot understand human intentions. This study aims at improving the robotic intelligence by training it to understand human intentions. Different from previous studies that recognizing human intentions from distinctive actions, this paper introduces a method to predict human intentions before a single action is completed. The experiment of throwing a ball towards designated targets are conducted to verify the effectiveness of the method. The proposed deep learning based method proves the feasibility of applying convolutional neural networks (CNN) under a novel circumstance. Experiment results show that the proposed CNN-vote method out competes three traditional machine learning techniques. In current context, the CNN-vote predictor achieves the highest testing accuracy with relatively less data needed.


Human-robot Interaction, Intentions Prediction, Convolutional Neural Networks