离线语义引导用于高效视觉-语言-动作策略蒸馏
Offline Semantic Guidance for Efficient Vision-Language-Action Policy Distillation
AI总结：
本文提出VLA-AD框架，利用视觉-语言模型作为离线语义监督者，将大规模VLA教师模型蒸馏为轻量学生策略，通过高阶语义指导提升效率与鲁棒性。
摘要：
大规模视觉-语言-动作（VLA）策略近期在机器人操作中表现出色，但其规模和推理成本仍是实时闭环控制的主要障碍。我们引入VLA-AD蒸馏框架，利用视觉-语言模型作为离线语义监督者，将大规模VLA教师模型转化为轻量学生策略。不同于仅依赖低层动作模仿，VLA-AD在教师提供的7自由度动作目标中加入高层语义指导，包括任务阶段锚点和多帧操作方向描述。这些辅助信号仅在训练时使用：在测试时，学生策略独立运行，无需VLA教师或VLM。我们在三个LIBERO基准测试套件上评估VLA-AD。使用OpenVLA-7B作为教师，我们的方法产生一个15800万参数的学生模型，模型大小减少44倍，同时与教师的平均相对差距仅为0.27%。生成的策略在RTX 4090上以12.5 Hz运行，比OpenVLA-7B快3.28倍。我们进一步表明，相同的语义蒸馏流程可泛化到不同的π_{0.5}-4B教师，其中学生在两个套件中优于教师，并在libero_goal上保持在0.53%以内。此外分析表明，阶段级监督和多帧方向线索使学生对噪声教师动作（如错误的高频夹具变化）更不敏感。总体而言，VLA-AD证明了从VLMs获得的离线语义指导可以显著提高VLA策略蒸馏的效率、鲁棒性和部署性。
英文摘要：
Billion-parameter Vision-Language-Action (VLA) policies have recently shown impressive performance in robotic manipulation, yet their size and inference cost remain major obstacles for real-time closed-loop control. We introduce \textbf{VLA-AD}, a distillation framework that uses a Vision-Language Model as an offline semantic supervisor to transfer large VLA teachers into lightweight student policies. Instead of relying only on low-level action imitation, VLA-AD augments teacher-provided 7-DoF action targets with high-level semantic guidance, including task phase anchors and multi-frame operating-direction descriptions. These auxiliary signals are used only during training: at test time, the student policy runs independently, with neither the VLA teacher nor the VLM required. We evaluate VLA-AD on three LIBERO benchmark suites. Using OpenVLA-7B as the teacher, our method produces a 158M-parameter student, yielding a $44\times$ reduction in model size while matching the teacher with only a $0.27\%$ average relative gap. The resulting policy runs at 12.5 Hz on an RTX 4090, achieving a $3.28\times$ inference speedup over OpenVLA-7B. We further show that the same semantic distillation pipeline generalizes to a different $π_{0.5}$-4B teacher, where the student outperforms the teacher on two suites and remains within $0.53\%$ on \texttt{libero\_goal}. Additional analysis indicates that phase-level supervision and multi-frame directional cues make the student less sensitive to noisy teacher actions, such as erroneous high-frequency gripper changes. Overall, VLA-AD demonstrates that offline semantic guidance from VLMs can substantially improve the efficiency, robustness, and deployability of VLA policy distillation.
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。
