Generative Adversarial Network for Imitation Learning from Single Demonstration

Baghdad Science Journal

Phan Xuan Tan¹  Tho Nguyen Duc¹  Chanh Minh Tran²  Eiji Kamioka² 

[1] School of Engineering and Science, Shibaura Institute of Technology, Japan.;School of Engineering and Science, Shibaura Institute of Technology, Japan;

关键词: Deep Learning, Few-shot Learning, Generative Adversarial Network, Imitation Learning, One-shot Learning;   

DOI  :  10.21123/bsj.2021.18.4(Suppl.).1350

来源: DOAJ

【 摘 要 】

Imitation learning is an effective method for training an autonomous agent to accomplish a task by imitating expert behaviors in their demonstrations. However, traditional imitation learning methods require a large number of expert demonstrations in order to learn a complex behavior. Such a disadvantage has limited the potential of imitation learning in complex tasks where the expert demonstrations are not sufficient. In order to address the problem, we propose a Generative Adversarial Network-based model which is designed to learn optimal policies using only a single demonstration. The proposed model is evaluated on two simulated tasks in comparison with other methods. The results show that our proposed model is capable of completing considered tasks despite the limitation in the number of expert demonstrations, which clearly indicate the potential of our model.

【 授权许可】

Unknown