1 paper
Ruifan Zuo, Guocheng Hu, Wanshui Gan +3
Cross-view spatial reasoning requires a model to align different viewpoints into a coherent spatial representation, yet this ability remains challenging for vision-language models…