Part of Advances in Neural Information Processing Systems 35 (NeurIPS 2022) Main Conference Track
Shaoshuai Shi, Li Jiang, Dengxin Dai, Bernt Schiele
Predicting multimodal future behavior of traffic participants is essential for robotic vehicles to make safe decisions. Existing works explore to directly predict future trajectories based on latent features or utilize dense goal candidates to identify agent's destinations, where the former strategy converges slowly since all motion modes are derived from the same feature while the latter strategy has efficiency issue since its performance highly relies on the density of goal candidates. In this paper, we propose the Motion TRansformer (MTR) framework that models motion prediction as the joint optimization of global intention localization and local movement refinement. Instead of using goal candidates, MTR incorporates spatial intention priors by adopting a small set of learnable motion query pairs. Each motion query pair takes charge of trajectory prediction and refinement for a specific motion mode, which stabilizes the training process and facilitates better multimodal predictions. Experiments show that MTR achieves state-of-the-art performance on both the marginal and joint motion prediction challenges, ranking 1st on the leaderbaords of Waymo Open Motion Dataset. Code will be available at https://github.com/sshaoshuai/MTR.