Image-Text Retrieval on ImageCoDe
5.9R@1SaMer (w/ FT)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SaMer (w/ FT)Architecture=Multi-vector, Backbone=ColPali, Token Compression (K)=64, Compression Strategy=Merging, Adapted=true2026.07 | 5.9 | 17 | 14.4 | |
| H-PoolArchitecture=Multi-vector, Backbone=ColPali, Token Compression (K)=64, Compression Strategy=Pooling2026.07 | 5.6 | 15.6 | 12.9 | |
| ColPaliArchitecture=Multi-vector, Backbone=ColPali, Compression=None2026.07 | 5.4 | 16 | 13.2 | |
| SaMer (w/o FT)Architecture=Multi-vector, Backbone=ColPali, Token Compression (K)=64, Compression Strategy=Merging, Adapted=false2026.07 | 5.2 | 17 | 14.1 | |
| SaMer (w/ FT)Architecture=Multi-vector, Backbone=ColQwen2, Token Compression (K)=64, Compression Strategy=Merging, Adapted=true2026.07 | 5.1 | 16.2 | 12.9 | |
| SAPArchitecture=Multi-vector, Backbone=ColQwen2, Token Compression (K)=64, Compression Strategy=Structural anchor pruning2026.07 | 4.8 | 13.7 | 11.8 | |
| SaMer (w/o FT)Architecture=Multi-vector, Backbone=ColQwen2, Token Compression (K)=64, Compression Strategy=Merging, Adapted=false2026.07 | 4.8 | 15.6 | 13 | |
| ColQwen2Architecture=Multi-vector, Backbone=ColQwen2, Compression=None2026.07 | 4.7 | 14.6 | 12.2 | |
| H-PoolArchitecture=Multi-vector, Backbone=ColQwen2, Token Compression (K)=64, Compression Strategy=Pooling2026.07 | 4.7 | 14.2 | 12.3 | |
| HPCArchitecture=Multi-vector, Backbone=ColQwen2, Token Compression (K)=64, Compression Strategy=Attention-based pruning2026.07 | 4.4 | 13.9 | 11.6 | |
| SAPArchitecture=Multi-vector, Backbone=ColPali, Token Compression (K)=64, Compression Strategy=Structural anchor pruning2026.07 | 3.8 | 12.5 | 10.1 | |
| HPCArchitecture=Multi-vector, Backbone=ColPali, Token Compression (K)=64, Compression Strategy=Attention-based pruning2026.07 | 3.6 | 11.3 | 9 | |
| SigLIP2Architecture=Single-vector2026.07 | 2.6 | 9.4 | 7.5 | |
| EVA-CLIPArchitecture=Single-vector2026.07 | 2.5 | 8.8 | 7.5 | |
| GMEArchitecture=VLM-based2026.07 | 2.3 | 9.1 | 7.7 | |
| OpenCLIPArchitecture=Single-vector2026.07 | 2.2 | 9.1 | 7.3 | |
| MetaCLIPArchitecture=Single-vector2026.07 | 2.1 | 7.6 | 6.3 | |
| DFN-CLIPArchitecture=Single-vector2026.07 | 1.9 | 9.8 | 7.7 | |
| VisRAGArchitecture=VLM-based2026.07 | 1.7 | 7 | 6 | |
| CLIPArchitecture=Single-vector2026.07 | 1.4 | 7 | 5.7 | |
| VLM2Vec-v2Architecture=VLM-based2026.07 | 1 | 5.1 | 4.4 |