BridgingtheGapbetweenObjectandImage-level RepresentationsforOpen-VocabularyDetection

Open in new window