Abstract: Open-vocabulary object detection (OVD) aims to detect novel object concepts by mining region-word correspondences from image-text pairs, yet current methods often produce false ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results