AutoSpeed: Annotation-Free Stage-Adaptive Motion Speed Learning for Robot Manipulation
2026-07-01 • Robotics
Robotics
AI summaryⓘ
The authors noticed that manipulation tasks have parts that are easier or harder, so the speed of movement should change accordingly. They created AutoSpeed, a method that helps existing robot vision-and-motion programs adjust how fast they predict and act, without needing extra labels. AutoSpeed tries different speeds for future actions and picks the best one based on prediction accuracy and timing. This lets robots go faster in simple parts and slower in tricky parts, improving both how quickly and how well tasks are done.
imitation learningvisuomotor policytemporal prediction horizontrajectory predictiondiscrete cosine transformspeed modulationrobot manipulationoptimizationmotion continuity
Authors
Qingda Hu, Ziheng Qiu, Jieru Zhao, Zhongxue Gan, Wenchao Ding
Abstract
Different stages of manipulation tasks exhibit varying levels of difficulty, suggesting stage-dependent motion speeds and temporal prediction horizons. However, existing IL-based visuomotor policies typically imitate the execution speed of expert demonstrations and operate with a fixed temporal prediction horizon, limiting flexibility and overall task throughput. In this paper, we introduce AutoSpeed, a model-agnostic learning framework that enables existing visuomotor policies to predict trajectories with stage-adaptive motion speeds, without requiring speed or stage annotations. We treat future trajectories at different speeds as candidate optimization targets, evaluate each candidate using a composite cost that trades off prediction error against prediction horizon, and optimize the policy toward the minimum-cost candidate. With a fixed-length action sequence, speed modulation adjusts the effective temporal prediction horizon: simple stages are executed faster with a longer prediction horizon, whereas complex stages are executed more slowly with a shorter prediction horizon. Specifically, we implement speed modulation in the frequency domain via the discrete cosine transform (DCT), which enables smooth, non-integer speed scaling and thus preserves motion continuity. Extensive evaluations show that AutoSpeed substantially reduces task execution time while also improving success rates. Under the AutoSpeed framework, the inferred motion speeds exhibit a strong correspondence with task stages.