arXiv · Computer Vision· Runtong Wu, Fei Teng, Di Wen, Guoqiang Zhao, Kunyu Peng, Kailun Yang·· 16 小时前AI 评分28
OmniAct3D：用基础几何与证据驱动推理做全景 3D 检测
OmniAct3D: Leveraging Foundation Geometry and Evidence-Grounded Reasoning for Panoramic 3D Detection
AI 导读
OmniAct3D 把基于视觉基础模型（VFM）的透视视角 3D 检测器适配到等距柱状投影（ERP）全景输入，提出 ERP-Ray Geometry Adapter（ERGA-Ray）、Visual-Action Reasoning Chain（VARC）与 Appearance-Guided Heading Expert（AGHE）三个模块。
来源：arXiv · Computer Vision · arxiv.org