Trajectory-Level Continuous Action Representation for Robotic Manipulation

2026-08-25

AI Summary

为解决机器人操作中动作表示与控制频率纠缠或依赖固定时间参数化的问题，提出CAT框架，通过编码动作轨迹为连续潜在标记，并加入频率感知位置编码，提高了不同控制频率下的表现。

Abstract

We propose CAT, a trajectory-level continuous action representation framework for robotic manipulation. Existing visuomotor systems often entangle action representation with control frequency or rely on fixed temporal parameterizations. This leads to representational redundancy at high sampling rates and limits the modeling of critical motion. CAT instead encodes action trajectories within a fixed real-time interval into a set of continuous latent tokens. To ensure temporal consistency across varying control frequencies, we further incorporate a frequency-aware positional encoding that establishs a shared temporal coordinate system. Trajectory-level regularization further stabilizes the latent representation. This approach prevents representation growth with timestep density and avoids reliance on predefined temporal parameterizations. Extensive system-level evaluations on LIBERO, MimicGen, and real-world long-horizon manipulation tasks demonstrate that CAT-based policies consistently outperform both competitive VQ-based and continuous visuomotor baselines under matched training settings. Across various model backbones and control frequencies, CAT consistently improves success rates. These results highlight the advantages of trajectory-level continuous action modeling for scalable robotic manipulation across varying control rates.

Problem

• action representation
• control frequency
• temporal parameterization
• redundancy

Innovation

• trajectory-level continuous action representation
• frequency-aware positional encoding
• latent tokens