LiMA: Bridging Long-term Imagination to Real-time Dexterous Manipulation via Asynchronous Diffusion

2026-09-23

AI Summary

为解决灵巧操作中长期预见与实时反应控制的时序错位问题，提出LiMA框架，通过异步双系统生成方式解耦意图规划与反应执行。

Abstract

Dexterous manipulation demands long-term foresight and rapid reactive control. Vision-Language-Action (VLA) models, while proficient in high-level reasoning, often lack a fine-grained understanding of physical dynamics and spatial perception. Conversely, World-Action Models (WAMs) typically suffer from high inference latency due to iterative generation. These deficiencies result in a critical temporal misalignment where the model's intent fails to adapt to rapid physical contact changes. To overcome this fundamental bottleneck, we propose LiMA, an asynchronous dual-system generative framework that systematically decouples intent planning from reactive execution. LiMA organizes computation into a multi-scale hierarchy: a slow system handles sparse long-horizon spatiotemporal intent generation, while a fast system focuses on dense high-frequency motion refinement. To align sparse intent predictions with dense action trajectories, we introduce a Latent Schrödinger Bridge Coupling mechanism that formulates refinement as an entropy-regularized probabilistic transport process. LiMA reduces inference latency by 45.8% compared with Cosmos-Policy via asynchronous decoupling. Evaluated across six bimanual dexterous manipulation tasks spanning multiple horizons, LiMA achieves an overall success rate of 70.8% and an average subtask success rate of 78.9%, while maintaining performance in unseen scenarios. The project website is available at https://ccdcs.github.io/LiMA_repo/

Problem

• Dexterous Manipulation
• Long-term Foresight
• Reactive Control
• Inference Latency
• Temporal Misalignment

Innovation

• asynchronous dual-system
• Latent Schrödinger Bridge Coupling
• multi-scale hierarchy
• intent planning and reactive execution decoupling