[121] Open-domain Visual Entity Recognition: Towards Recognizing Millions of Wikipedia Entitiesmultimodal CLIP 2023Q1 retrieval