Q-GeoMem: Question-Guided Geometric Memory for Video Spatial Reasoning 文章

ArXiv CS.CV2026-05-27NEWSen作者: Xianqiang Gao, Qizhi Chen, Delin Qu, Haoming Song, Zhigang Wang, Bin Zhao, Dong Wang, Xuelong Li

查看原文 →

Q-GeoMem: Question-Guided Geometric Memory for Video Spatial Reasoning · 相关技术

相关技术

ORM ODE language model 远程代码执行（RCE）spatial reasoning divide-and-conquer partitioning cameras UCT Token SPA SEM Referring expression comprehension (REC)Parts-of-Speech (POS) tags GEO For Effort Metric Attention DiT Dense ANN