الملخص
Pretrained robot policies offer strong manipulation skills but are typically limited to single-agent settings, where a robot acts in isolation. In this work, we study how to adapt pretrained single-agent diffusion policies to multi-agent settings using minimal collaborative data, co-optimizing for two key objectives: high coordination performance and single-agent skill retention. To this end, we introduce ALTER, an adaptation method for coordination on demand: the adapted policy coordinates with other robots when deployed in a team while remaining capable of acting independently when operating alone. Execution is decentralized: each robot acts only on its own visual observations, without explicit inter-agent communication. Our method trains a coordination head that predicts a residual denoiser to transform single-agent behavior into coordinated multi-agent behavior when necessary while also preserving single-agent capabilities. To preserve single-agent capabilities, we augment a small number of collaborative demonstrations with self-distilled data generated by the base policy during training of the residual denoiser. In simulation, ALTER achieves higher coordination success over our baselines while retaining much higher source-skill retention. In our hardware experiments, we find similar trends where ALTER better co-optimizes for coordination success and single-agent skill retention than the baselines.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Dong, D., Bhatt, M., Shrivastava, A., Peters, L., & Mehr, N. (2026). Residual Denoising Enables Sample-Efficient Multi-Agent Coordination on Demand. https://omanscience.com/ar/articles/residual-denoising-enables-sample-efficient-multi-agent-coordination-on-demand
MLA 9
Dong, Dayi, et al. "Residual Denoising Enables Sample-Efficient Multi-Agent Coordination on Demand." https://omanscience.com/ar/articles/residual-denoising-enables-sample-efficient-multi-agent-coordination-on-demand.
شيكاغو (المؤلف–التاريخ)
Dong, Dayi, Maulik Bhatt, Aayushi Shrivastava, Lasse Peters, and Negar Mehr. 2026. "Residual Denoising Enables Sample-Efficient Multi-Agent Coordination on Demand." https://omanscience.com/ar/articles/residual-denoising-enables-sample-efficient-multi-agent-coordination-on-demand.
هارفارد
Dong, D., Bhatt, M., Shrivastava, A., Peters, L. and Mehr, N. (2026) 'Residual Denoising Enables Sample-Efficient Multi-Agent Coordination on Demand', Available at: https://omanscience.com/ar/articles/residual-denoising-enables-sample-efficient-multi-agent-coordination-on-demand.
فانكوفر
Dong D, Bhatt M, Shrivastava A, Peters L, Mehr N. Residual Denoising Enables Sample-Efficient Multi-Agent Coordination on Demand. https://omanscience.com/ar/articles/residual-denoising-enables-sample-efficient-multi-agent-coordination-on-demand
IEEE
D. Dong, M. Bhatt, A. Shrivastava, L. Peters, and N. Mehr, "Residual Denoising Enables Sample-Efficient Multi-Agent Coordination on Demand," https://omanscience.com/ar/articles/residual-denoising-enables-sample-efficient-multi-agent-coordination-on-demand.