EUVIMEDEUVIMED entdecken
Crossref

UniMM: A Unified Mixture Model Framework for Multi-Agent Simulation

Longzhong Lin, Xuewu Lin, Kechun Xu, Haojian Lu, Lichao Huang, Rong Xiong, Yue Wang

IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-10 · 48 (10) · 12601-12618

DOI: 10.1109/tpami.2026.3700402 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-10. Verfasst von Longzhong Lin und 6 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 59 Literaturangaben. Gefördert von: National Nature Science Foundation, Zhejiang Provincial Natural Science Foundation of China, Key R&D Program of Hangzhou.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Simulation plays a crucial role in assessing autonomous driving systems, where the generation of realistic multi-agent behaviors is a key aspect. In multi-agent simulation, the primary challenges include behavioral multimodality and closed-loop distributional shifts. In this study, we formulate a unified mixture model (UniMM) framework for generating multimodal agent behaviors, which can cover the mainstream methods including regression-based mixture models and discrete NTP models. Furthermore, we introduce a closed-loop sample generation approach tailored for mixture models to mitigate distributional shifts. Within the UniMM framework, we recognize critical configurations from both the model and data perspectives. We conduct a systematic examination of various model configurations, and comprehensively characterize their effects. Moreover, our investigation into the data configuration highlights the pivotal role of closed-loop samples in achieving realistic simulations. To extend the benefits of closed-loop samples across a broader range of mixture models, we further introduce a temporal disentanglement-and-alignment mechanism to address the shortcut learning and off-policy learning issues. Leveraging insights from our exploration, the distinct variants proposed within the UniMM framework, including discrete, anchor-free, and anchor-based models, all achieve state-of-the-art performance on the WOSAC benchmark.

Förderung

  • National Nature Science Foundation
  • Zhejiang Provincial Natural Science Foundation of China
  • Key R&D Program of Hangzhou

Zitieren

APA

Lin, L., Lin, X., Xu, K., Lu, H., Huang, L., Xiong, R., & Wang, Y. (2026). UniMM: A Unified Mixture Model Framework for Multi-Agent Simulation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(10), 12601-12618. https://doi.org/10.1109/tpami.2026.3700402

Vancouver

Lin L, Lin X, Xu K, Lu H, Huang L, Xiong R, et al. UniMM: A Unified Mixture Model Framework for Multi-Agent Simulation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(10):12601-12618. doi:10.1109/tpami.2026.3700402

BibTeX
@article{Lin2026,
  author = {Lin, Longzhong and Lin, Xuewu and Xu, Kechun and Lu, Haojian and Huang, Lichao and Xiong, Rong and Wang, Yue},
  title = {UniMM: A Unified Mixture Model Framework for Multi-Agent Simulation},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2026},
  volume = {48},
  number = {10},
  pages = {12601--12618},
  doi = {10.1109/tpami.2026.3700402}
}
Kostenlos starten →