MotionLM: Multi-Agent Motion Forecasting as Language Modeling

Autor:	Seff, Ari, Cera, Brian, Chen, Dian, Ng, Mason, Zhou, Aurick, Nayakanti, Nigamaa, Refaat, Khaled S., Al-Rfou, Rami, Sapp, Benjamin
Rok vydání:	2023
Předmět:	Computer Science - Computer Vision and Pattern Recognition Computer Science - Artificial Intelligence Computer Science - Machine Learning Computer Science - Robotics
Druh dokumentu:	Working Paper
Popis:	Reliable forecasting of the future behavior of road agents is a critical component to safe planning in autonomous vehicles. Here, we represent continuous trajectories as sequences of discrete motion tokens and cast multi-agent motion prediction as a language modeling task over this domain. Our model, MotionLM, provides several advantages: First, it does not require anchors or explicit latent variable optimization to learn multimodal distributions. Instead, we leverage a single standard language modeling objective, maximizing the average log probability over sequence tokens. Second, our approach bypasses post-hoc interaction heuristics where individual agent trajectory generation is conducted prior to interactive scoring. Instead, MotionLM produces joint distributions over interactive agent futures in a single autoregressive decoding process. In addition, the model's sequential factorization enables temporally causal conditional rollouts. The proposed approach establishes new state-of-the-art performance for multi-agent motion prediction on the Waymo Open Motion Dataset, ranking 1st on the interactive challenge leaderboard. Comment: To appear at the International Conference on Computer Vision (ICCV) 2023
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2309.16534 Zobrazit plný text záznamu View this record from Arxiv