Abstract
Multi-agent trajectory prediction plays a crucial role in robotics and autonomous driving. The current mainstream research focuses on how to achieve accurate prediction on one large dataset. However, whether the multi-agent trajectory prediction model can be trained with sequential datasets, i.e. continual learning settings, remains a question. Can the current prediction methods avoid catastrophic forgetting Can we utilize the continual learning strategy in multi-agent trajectory prediction application Motivated by the rehearsal methods in continual learning literature, we propose a multi-agent interaction behavior prediction framework with graph-neural-network-based conditional generative memory system to mitigate catastrophic forgetting. To the best of our knowledge, this work is the first attempt to study the continual learning problem in multi-agent interaction behavior prediction problems. We empirically show that several approaches in literature indeed suffer from catastrophic forgetting, and our approach succeeds in maintaining a low prediction error when data comes in a sequential way. We also conduct an ablation analysis to show the effectiveness and necessity of our proposed conditional generative memory system.