Rolling-WAM：滑动窗口式去噪让机器人重规划提速4.5倍
Rolling-WAM: World Action Models with Rolling Imagination
机器人控制的延迟问题有新解法了，Rolling-WAM 把去噪拆到多个重规划周期里做，实测比标准方法快4.5倍，还在真实 Unitree G1 上跑通了。
Rolling-WAM 是一种针对世界动作模型（WAM）的去噪优化方法，解决了每次重规划都要完整完成视频-动作联合去噪导致延迟高的问题。它维护一个处于不同噪声等级的视频-动作块滑动窗口，每步只完整去噪即将执行的动作块，较远的未来块则部分去噪并随窗口推进继续处理。在 LIBERO、RoboTwin 基准和真实 Unitree G1 人形机器人上的评估显示其操作性能与标准方法相当，同时稳态重规划速度提升 4.5 倍。论文已发布于 arXiv。
Rolling-WAM: World Action Models with Rolling Imagination
World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation. However, completing the joint video-action denoising process at each replanning cycle incurs substantial latency, delaying action updates and limiting closed-loop responsiveness. We present Rolling-WAM, a formulation that distributes joint denoising across successive replanning cycles. Our method maintains a sliding window of video-action chunks at staggered noise levels. At each step, a rolling noise schedule fully denoises the imminent action chunk for execution, while partially refining farther-future chunks. As the window advances with new camera observations, the retained future chunks continue their denoising process. This distributes the computational cost over time while carrying an evolving visual-action context across chunk boundaries. Evaluations on LIBERO, RoboTwin, and a real-world Unitree G1 humanoid show that Rolling-WAM achieves competitive manipulation performance. By removing the need to denoise the entire prediction horizon from scratch, it delivers a 4.5x steady-state replanning speedup over standard joint WAMs.