![[2506.08822] FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistency](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2506.08822] FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistency
Summary
The paper presents FreqPolicy, a novel flow-based visuomotor policy that enhances efficiency in robotic manipulation by imposing frequency consistency constraints, enabling high-quality one-step action generation.
Why It Matters
This research addresses the critical challenge of high inference costs in robotic systems by introducing a method that maintains temporal coherence in action generation. The findings could significantly improve real-time robotic applications, making them more efficient and effective.
Key Takeaways
- FreqPolicy improves efficiency in robotic manipulation tasks.
- The approach enforces frequency consistency for better temporal coherence.
- Demonstrated superiority over existing one-step action generators across multiple benchmarks.
Computer Science > Robotics arXiv:2506.08822 (cs) [Submitted on 10 Jun 2025 (v1), last revised 18 Feb 2026 (this version, v2)] Title:FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistency Authors:Yifei Su, Ning Liu, Dong Chen, Zhen Zhao, Kun Wu, Meng Li, Zhiyuan Xu, Zhengping Che, Jian Tang View a PDF of the paper titled FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistency, by Yifei Su and 8 other authors View PDF HTML (experimental) Abstract:Generative modeling-based visuomotor policies have been widely adopted in robotic manipulation, attributed to their ability to model multimodal action distributions. However, the high inference cost of multi-step sampling limits its applicability in real-time robotic systems. Existing approaches accelerate sampling in generative modeling-based visuomotor policies by adapting techniques originally developed to speed up image generation. However, a major distinction exists: image generation typically produces independent samples without temporal dependencies, while robotic manipulation requires generating action trajectories with continuity and temporal coherence. To this end, we propose FreqPolicy, a novel approach that first imposes frequency consistency constraints on flow-based visuomotor policies. Our work enables the action model to capture temporal structure effectively while supporting efficient, high-quality one-step action generation. Concretely, we introduce a frequency consistenc...
