arXiv · Computer Vision· Jingyu Guo, Jiaxin Huang, Chengxing Lin, Xiaoyu Luo, Cheng Wen, Shanshan Ye, Tingjin Chu, Shaoli Huang, Mingming Gong·· 3 hr agoAI score32
GuidedNav：通过预行动注意力与空间引导塑造视觉-语言导航表征
GuidedNav: Shaping Vision-Language Navigation Representations through Pre-action Attention and Spatial Guidance
AI brief
GuidedNav 是一个统一框架，通过训练期引导塑造视觉-语言导航（VLN）表征：子任务与地标引导将预行动注意力对准当前指令段与相关视觉区域，空间引导则对齐几何特征以捕捉长程上下文与局部空间结构。
Source: arXiv · Computer Vision · arxiv.org