Less Context, Better Geometry: Masked Geometric Encoder for Robust 3D Foundation Models

作者：Zhimin Shao, Xijun Liu, Zhaoliang Zhang, Yutao Tang, Abhay Yadav, Rama Chellappa, Cheng Peng · 单位：Department of Electrical and Computer Engineering, Johns Hopkins University, Department of Data Science, University of Virginia · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-10-06