Fetching the paper…

COTS: Collaborative Two-Stream Vision-Language Pre-Training Model for Cross-Modal Retrieval · Around