UniMotion: A Unified Motion Framework for Simulation, Prediction and Planning
Nan Song, Junzhe Jiang, Jingyu Li, Xiatian Zhu, Li Zhang · 2026
Read UniMotion: A Unified Motion Framework for Simulation, Prediction and Planning by Nan Song, Junzhe Jiang, Jingyu Li, Xiatian Zhu, Li Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-00566.txt.
Description
Nan Song and colleagues unify motion simulation, prediction, and planning for autonomous driving, tasks usually handled apart despite shared needs. Separate designs hinder cross-task generalisation and scale, and miss mutual benefits between tasks. Their UniMotion framework, built on a decoder-only Transformer, uses dedicated interaction modes and tailored training to serve all three tasks at once. Joint training aids generalisation and task integration, and targeted fine-tuning lifts UniMotion to state-of-the-art results across motion tasks on the Waymo Open Motion Dataset.
Catalog record
| CREATOR | Nan Song, Junzhe Jiang, Jingyu Li, Xiatian Zhu, Li Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40197 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.00566