GaussTR: Foundation Model-Aligned Gaussian Transformer for Self-Supervised 3D Spatial Understanding figure
AlphaXiv 中文概览(可滚动查看)