Abstract: 3D visual grounding is a critical skill for household robots, enabling them to navigate, manipulate objects, and answer questions based on their environment. While existing approaches often ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback