Preprint
Oct 2026
EgoAlign: Situated 3D Understanding without Situation Descriptions
Results show that EgoAlign establishes agent-centric spatial context directly from numerical pose without requiring textual situation descriptions, and achieves the strongest performance on SQA3D, Real-3DQA, and MSQA among compared methods while remaining highly efficient.
Yi-Yang Li, Tian-Yi Ma, Ze-Hong Wang et al.
· 0 citations