CORTEXA
← Browse

Xuanran Dong

1 paper indexed

arxivcs.CVcs.AI2026-07-16

VTM-Nav: Hierarchical Visual-Topological Memory for Cross-Episode Object-Goal Navigation

Xiaoran Xu, Yupeng Wu, Tianyu Xue, Yifan Xu, Xuanran Dong, Xiaoshan Yang, et al.

Object-goal navigation requires an embodied agent to locate and reach an instance of a specified object category in an indoor environment. Recent training-free approaches leverage vision-language models (VLMs) for open-vocabulary semantic reasoning, but are typically evaluated un…

View free PDFSource page