1 paper · 2 filters
Qi Chen, Changli Wu, Jiayi Ji +2
Current 3D visual grounding tasks only process sentence level detection or segmentation, which critically fails to leverage the rich compositional contextual reasonings within natu…