Aligning Bag of Regions for Open-Vocabulary Object DetectionShare on Twitter Facebook LinkedIn Previous Next