arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.07904cs.LGcs.RO
收藏

DyQ-VLA: 时序动态感知的视觉-语言-动作模型量化

DyQ-VLA: Temporal-Dynamic-Aware Quantization for Embodied Vision-Language-Action Models

Zihao Zheng, Hangyu Cao, Chibang Tao, Zhihao Mao, Lingyue Zhang, Maoliang Li, Jiayu Chen, Hailong Zou, Xuanzhe Liu, Donggang Cao, Hong Mei, Xiang Chen

AI总结:

本文提出DyQ-VLA,通过动态量化框架解决视觉-语言-动作模型在时序动态敏感性和实时分配上的问题,实现内存占用降低30.9%且性能保持99.5%,在仿真和现实场景中分别获得1.49倍和1.43倍的速度提升。

摘要:

视觉-语言-动作(VLA)模型在具身智能中占主导地位,但受推理开销限制。尽管模型量化缓解了边缘部署的瓶颈,但静态量化方法仍无法满足VLA需求,因存在两个关键挑战:(1)时序动态敏感性,固定精度因忽略阶段变化的误差容忍度而浪费资源;(2)实时分配,识别实时敏感性以指导位宽分配仍无解。为解决这些问题,我们提出DyQ-VLA,一种专为VLA设计的动态量化框架。具体而言,一个敏感性感知的切换策略利用实时运动学代理触发位宽切换,而一个运动学引导模块动态分配最优位宽。实验表明,DyQ-VLA仅需原始内存占用的30.9%,同时保持99.5%的原始性能,实现1.49倍的仿真加速和高达1.43倍的现实场景加速。

英文摘要:

Vision-language-action (VLA) models achieve favorable task performance, yet runtime errors in closed-loop execution evolve with alternating updates of actions and observations. Existing VLA quantization methods mainly trade off inference speed and model performance while rarely investigating how quantization alters runtime errors. By comparing error distributions between full-precision and quantized models, we identify distinct evolutionary patterns for the two types of errors: quantization enlarges the variance of translation error distributions and increases their dispersion, whereas the distribution center of rotation errors gradually shifts across execution steps, demonstrating cumulative drift. Motivated by this observation, we rethink the optimal quantization strategy for VLA models and propose \textit{DyQ-VLA}, a runtime-error-aware quantization framework, which dynamically selects activation precision according to execution steps and tracks as well as compensates rotation errors via accumulated quantization residuals. Corresponding operators and runtime adaptation strategies are devised within the framework to enable dynamic-precision execution. Experiments show that \textit{DyQ-VLA} achieves a 1.88 to 1.93 inference speedups while maintaining comparable or higher average task success rates. Moreover, it reduces mean execution steps by 17.7% to 19.8%. Our code is here: https://anonymous.4open.science/r/DyQ-VLA-7F51/.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑