DH-Set: Improving Vision-Language Alignment with Diverse and Hybrid Set-Embeddings Learning
Open paper
Benchmark
Model
Rank
Results
cross-modal-retrieval-on-flickr30k
DH-Set (CLIP-based ViT-large-384 + BERT-large)
–
Image-to-text R@1: 96.3
Rank counts only results with a code link.