Haibing Ren1, Hu Chen1, Zhongxuan Liu1, Yimin Zhang
10:00 - 10:30 | Mon 25 Sep | Ballroom Foyer | MoAmPo.36
This paper proposed reinforcement learning with function module network method for human robot interaction. It divides the whole Q networks into several functional modules which may easily transfer the trained model to other tasks. With the new reward and loss function, local navigation and human robot interaction are solved simultaneously.