← Back to Benchmarks
simmediumroboticsmetric · varies

MotionLM: Multi-Agent Motion Forecasting as Language Modeling

Description

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

Source

http://arxiv.org/abs/2309.16534v1