RN-50 (NegCLIP, CC-12M)

Reported on 2 benchmarks across 1 task · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision2 results

Image RetrievalonCREPE (Compositional REPresentation Evaluation)
Recall@1 (HN-Atom, UC)· 2023-05-23
41.4
best: 47.86 (ViT-L-14 (LAION400M))
Coarse-to-Fine Contrastive Learning in Image-Text-Graph Space for Improved Vision-Language Compositionality arXiv:2305.13812
Image RetrievalonCREPE (Compositional REPresentation Evaluation)
Recall@1 (HN-Comp, UC)· 2023-05-23
82
best: 92.6 (RN-50 (MosaiCLIP, CC-12M))
Coarse-to-Fine Contrastive Learning in Image-Text-Graph Space for Improved Vision-Language Compositionality arXiv:2305.13812