FocalPolicy: Frequency-Optimized Chunking and Locally Anchored Flow Matching for Coherent Visuomotor Policy

2026-05-15

AI Summary

Existing visuomotor policies struggle to simultaneously achieve local precision and global coherence when generating long-horizon action trajectories, often resulting in discontinuities across action segments. To address this, this work proposes FocalPolicy, which innovatively integrates frequency-aware chunking with a locally anchored flow matching mechanism. Specifically, a forward-looking composite objective function jointly optimizes temporal alignment of local actions and structural consistency across chunks in the frequency domain. Additionally, a locally anchored sampling strategy is introduced to enhance the propagation efficiency of target signals within the flow matching framework. Experiments demonstrate that FocalPolicy significantly outperforms current methods across multiple visuomotor tasks, and its components can be effectively transferred to other baselines, confirming both its generality and efficacy.

Abstract

Visuomotor policies aim to learn complex manipulation tasks from expert demonstrations. However, generating smooth and coherent trajectories remains challenging, as it requires balancing proximal precision with distal foresight. Existing approaches typically focus on optimizing intra-chunk action distributions, often neglecting the inter-chunk coherence. Consequently, inter-chunk discontinuities significantly impede the learning of coherent long-horizon actions. To overcome this limitation and achieve a synergetic balance between precision and foresight, we propose FocalPolicy, a foresight-aware visuomotor policy that combines Frequency-Optimized Chunking with Locally Anchored flow matching. We introduce a foresight composite objective that supervises time-domain alignment within the proximal actions while regularizing frequency-domain structure over multiple future action chunks to improve cross-chunk coherence. To efficiently learn complex action distributions, we design locally anchored campling to enhance target signal propagation efficiency during consistency flow matching training. Extensive experiments demonstrate that FocalPolicy outperforms existing approaches and confirm the generalizability of our modules to other baselines. Project website: https://focalpolicy.github.io/

Problem

• visuomotor policy
• action coherence
• chunking
• long-horizon manipulation
• trajectory smoothness

Innovation

• Frequency-Optimized Chunking
• Locally Anchored Flow Matching
• Visuomotor Policy
• Cross-chunk Coherence
• Foresight-aware Learning