DH-Set: Improving Vision-Language Alignment with Diverse and Hybrid Set-Embeddings Learning.

Kun Zhang, Jingyu Li, Zhe Li, S. Kevin Zhou

Browse the full CVPR paper archive.