Abstract: Visual grounding aims to use a natural language expression to find specific objects in an image, whether in a bounding box or a segmentation mask. The vision research community has ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback