CORTEXA
← Browse

Xiaoran Xu

2 papers 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
arxivcs.CV2026-07-14

Hy-Embodied-VLM-1.0: Efficient Physical-World Agents

Ziyi Wang, Xumin Yu, Yongming Rao, Yonggen Ling, Yunheng Li, Oran Wang, et al.

Building capable embodied agents requires not only multimodal perception and understanding, but also agentic capabilities for reasoning about actions, adapting to evolving situations, and interacting with the physical world. In this report, we introduce Hy-Embodied-VLM-1.0, an ef…

View free PDFSource page