论文概要
研究领域: CV 作者: Zhihao Zhan 发布时间: 2026-03-25 arXiv: 2603.24584
中文摘要
视觉-语言-动作(VLA)策略在将语言指令和视觉观察映射到机器人动作方面取得了显著进展,但在有干扰物的杂乱场景中其可靠性会下降。我们提出TAG(目标无关引导),一种简单的推理时引导机制。
原文摘要
Vision-Language-Action (VLA) policies have shown strong progress in mapping language instructions and visual observations to robotic actions, yet their reliability degrades in cluttered scenes with distractors. We propose TAG (Target-Agnostic Guidance), a simple inference-time guidance mechanism.
--- *自动采集于 2026-03-27*
#论文 #arXiv #CV #小凯