ICML 2026 | Breaking 3DGS Photometric Ambiguity: Beihang/NUS Propose AmbiSuR for High-Fidelity 3D Geometry Reconstruction
机器之心86 字 (约 1 分钟)
65
A joint research team from Beihang University and National University of Singapore presents AmbiSuR at ICML 2026, proposing an innovative solution for the photometric ambiguity problem in 3D Gaussian Splatting, with potential to enhance high-fidelity 3D geometry reconstruction.
入选理由：AmbiSuR方法针对3DGS光度多义性瓶颈提出解决方案
FeaturedArticle#3DGS#3D Reconstruction#Computer Vision#ICML#Academic Paper中文