arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.22982cs.RO
收藏

将协作动态蒸馏到潜在空间以实现分散式多智能体操作中的隐式协调

Distilling Collaborative Dynamics into Latent Space for Implicit Coordination in Decentralized Multi-Agent Manipulation

Korea Advanced Institute of Science and Technology (KAIST)(韩国科学技术院)

Chanyoung Park, Minsung Yoon, Andrew Jeong, Sung-eui Yoon

AI总结:

提出CLS-DP分散式多智能体框架,通过将多智能体动态蒸馏到潜在空间,在部分可观测下实现隐式协调,无需全局视图或通信,在RoboFactory任务中平均成功率38%,优于集中式基线20%。

摘要:

多臂操作需要精确的时空协调,然而许多集中式方法随着团队规模增大而扩展性差。为了解决这一问题,我们提出了CLS-DP,一种分散式多智能体框架,能够在部分可观测性下实现隐式协调,无需共享全局视图、显式状态信息或智能体间通信。在集中式训练与分散式执行(CTDE)范式下,CLS-DP将特权多智能体动态蒸馏到潜在空间。在部署时,每个智能体从其局部RGB观测和共享任务指令中推断出协作潜在向量;然后,它将该潜在向量作为扩散去噪过程的条件。这种设计实现了隐式协调,且每个智能体的成本与团队规模无关。在跨越两到四个智能体的六个RoboFactory基准任务中,CLS-DP达到了38%的平均成功率,优于最佳集中式基线(20%)和没有协作潜在向量的分散式消融模型(9%)。在所有智能体配置中,它还保持了优越的参数效率。归因图显示,以协作潜在向量为条件的智能体在整个执行过程中,对其自身和队友的关节和夹爪都给予了高度关注。这表明,学习到的潜在向量有效地从局部观测中编码了协作动态,从而在部分可观测的现实场景中促进了隐式协调。

英文摘要:

Multi-arm manipulation demands precise spatiotemporal coordination, yet many centralized approaches scale poorly as team size increases. To address this, we propose CLS-DP, a decentralized multi-agent framework that enables implicit coordination under partial observability without shared global views, explicit state information, or inter-agent communication. Under the centralized training and decentralized execution (CTDE) paradigm, CLS-DP distills privileged multi-agent dynamics into a latent space. At deployment, each agent infers a collaborative latent from its local RGB observation and a shared task instruction; it then conditions the diffusion denoising process on this latent. This design enables implicit coordination with a per-agent cost independent of team size. Across six RoboFactory benchmark tasks spanning two to four agents, CLS-DP achieves a 38% mean success rate, outperforming the best centralized baseline (20%) and a decentralized ablation without the collaborative latent (9%). It also maintains superior parameter efficiency across all agent configurations. Attribution maps show that an agent conditioned on the collaborative latent places high attribution on the joints and grippers of both itself and its teammates throughout execution. This suggests that the learned latent efficiently encodes collaborative dynamics from local observation, which facilitates implicit coordination in realistic settings characterized by partial observability.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑