When to Switch: Reliable Action-Chunk Extension for Vision-Language-Action Models

📄 arXiv: 2610.05719v1 📥 PDF

作者: Seonghoon Yu, Dongwon Kim, HyungRok Jung, Yoonjae Baek, Byung-kwan Lee, Suha Kwak, Jeany Son

分类: cs.RO

发布日期: 2026-10-05

备注: Pre-print

🔗 代码/项目: GITHUB


💡 一句话要点

提出RACE框架以解决VLA模型中的动作切换不可靠问题

🎯 匹配领域: 支柱一:机器人控制 (Robot Control) 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 视觉-语言-动作 机器人操作 动作块扩展 子技能切换 深度学习

📋 核心要点

  1. 现有VLA模型在推理时需要频繁暂停,导致动作不连贯和任务完成时间延长。
  2. 提出RACE框架,通过预测动作切换时机来减少子技能过渡处的错误,提高长动作块的执行可靠性。
  3. 实验表明,RACE在模拟和实际机器人任务中均显著提高了成功率,并减少了停顿时间。

📝 摘要(中文)

视觉-语言-动作(VLA)模型作为机器人操作的统一策略,因推理成本高导致机器人在策略调用之间需要暂停,造成动作不连贯和任务完成时间延长。通过扩展动作块可以减少策略调用次数,但长时间块的执行可靠性下降。本文分析了长块中的动作错误,发现错误主要集中在操作子技能之间的过渡时刻。为此,提出RACE(可靠动作块扩展)框架,通过辅助的单步去噪过程预测过渡时机,并基于此条件生成动作。RACE显著降低了子技能过渡处的错误,提升了长动作块的可靠执行。实验结果表明,RACE在模拟基准测试中超越了同块长度的微调方法,并在2倍长的块中成功率超过了最新的高效VLA,在4倍长的块中仍保持竞争力。实际机器人实验中,RACE使用4倍长的块,减少了约5倍的停顿时间,同时成功率高于相同块长度的微调。

🔬 方法详解

问题定义:本文旨在解决视觉-语言-动作(VLA)模型在执行长动作块时的可靠性问题。现有方法在长块执行中容易出现错误,尤其是在操作子技能之间的过渡时刻,导致执行不稳定。

核心思路:RACE框架的核心思想是通过辅助的单步去噪过程来预测子技能的切换时机,并在动作生成时以此为条件,从而减少过渡处的错误,提高长动作块的执行可靠性。

技术框架:RACE框架主要包括两个阶段:第一阶段是通过去噪网络预测切换时机,第二阶段是基于预测的切换时机生成动作。这一流程确保了动作生成与切换时机的紧密结合。

关键创新:RACE的主要创新在于引入了切换时机的预测机制,这一机制有效降低了子技能过渡处的错误率,与传统的微调方法相比,能够在更长的动作块中保持高成功率。

关键设计:在设计中,RACE使用了特定的损失函数来优化切换时机的预测精度,并采用了改进的网络结构以增强模型的学习能力。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

在模拟基准测试中,RACE在相同块长度下超越了微调方法,并在2倍长的块中成功率超过了最新的高效VLA,4倍长的块中仍保持竞争力。在实际机器人实验中,RACE使用4倍长的块,减少了约5倍的停顿时间,同时成功率高于相同块长度的微调。

🎯 应用场景

该研究的潜在应用领域包括机器人操作、自动化制造和智能家居等场景。通过提高VLA模型的执行可靠性,RACE框架能够显著提升机器人在复杂任务中的表现,具有重要的实际价值和广泛的应用前景。

📄 摘要(原文)

Vision-Language-Action (VLA) models serve as unified policies for robotic manipulation, yet their expensive inference forces robots to pause between policy calls, resulting in stop-and-go execution that interrupts smooth motion and prolongs task completion. Extending the action chunk reduces policy calls and hence these pauses, but predicting farther into the future makes long-chunk execution unreliable. To understand where this unreliability arises, we analyze action errors within long chunks and find that they concentrate around transitions between manipulation subskills, growing sharply with chunk length. This suggests the importance of transition timing, i.e., when to switch subskills within a chunk. Motivated by this observation, we introduce RACE (Reliable Action-Chunk Extension), a framework that predicts the transition timing from an auxiliary one-step denoising pass and conditions action generation on it. By learning and conditioning on transition timing, RACE reduces errors at subskill transitions and enables reliable execution of longer action chunks. Across simulation benchmarks, RACE outperforms fine-tuning at the same chunk length; with 2x longer chunks, it surpasses recent state-of-the-art and efficient VLAs in success rate, and with 4x longer chunks, it remains competitive. On a real robot, RACE uses 4x longer chunks, which reduces the idle time caused by stop-and-go execution by about 5x, while achieving a higher success rate than fine-tuning with the same chunk length. Code and a real-robot demo are available at https://github.com/Seonghoon-Yu/RACE-VLA