Fetching the paper…

Bridging the Gap between Object and Image-level Representations for Open-Vocabulary Detection · Around