Image-to-Text Retrieval on COCO-CN
87.4R@1ObjEmbed
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ObjEmbedParameters=4B2026.02 | 87.4 | — | — | — | — | |
| ObjEmbedParameters=2B2026.02 | 86 | — | — | — | — | |
| UME-R1Parameters=2B2026.02 | 83.6 | — | — | — | — | |
| FG-CLIP2Backbone=ViT-So/162026.02 | 83.2 | — | — | — | — | |
| Qwen3-VL-EmbeddingParameters=2B2026.02 | 83 | — | — | — | — | |
| GMEParameters=7B2026.02 | 81.7 | — | — | — | — | |
| M2-EncoderModel Size=10B, Protocol=Zero-shot2024.01 | 80.9 | 96.5 | 99.1 | 91.7 | — | |
| MetaCLIP2Backbone=ViT-H/142026.02 | 80.1 | — | — | — | — | |
| GMEParameters=2B2026.02 | 79.7 | — | — | — | — | |
| AltCLIPModel Size=0.8B, Protocol=Zero-shot2024.01 | 77.8 | 94.4 | 97.5 | 85 | — | |
| M2-EncoderModel Size=1B, Protocol=Zero-shot2024.01 | 74.6 | 94.2 | 98 | 89.3 | — | |
| VLM2Vec-V2Parameters=2B2026.02 | 74.4 | — | — | — | — | |
| Wukong_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 73.3 | 94 | 98 | 88.6 | — | |
| SigLIP2Backbone=ViT-So/162026.02 | 72 | — | — | — | — | |
| M2-EncoderModel Size=0.4B, Protocol=Zero-shot2024.01 | 71 | 93.7 | 97.9 | 87.8 | — | |
| FILIP_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 69.1 | 91.3 | 96.9 | 86.5 | — | |
| ERNIE-ViL 2.0Setting=Zero-shot2022.09 | 69.1 | 92.9 | 97.1 | — | 86.1 | |
| ERNIE-ViL 2.0Protocol=Zero-shot2024.01 | 69.1 | 92.9 | 97.1 | 86.1 | — | |
| CLIP_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 68.7 | 93.6 | 97.5 | 86.6 | — | |
| CLIP_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 68.3 | 93 | 97.3 | 86.2 | — | |
| FILIP_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 68.3 | 93.9 | 97.1 | 86.7 | — | |
| CLIP_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 68 | 92.8 | 97.3 | 85.4 | — | |
| VLM2VecParameters=2B2026.02 | 67.9 | — | — | — | — | |
| Wukong_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 67.4 | 92.4 | 97.5 | 85.5 | — | |
| Wukong_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 65.8 | 90.3 | 96.6 | 84.6 | — | |
| R2D2Setting=Zero-shot2022.09 | 63.3 | 89.3 | 95.7 | — | 80.5 | |
| R2D2Backbone=ViT-L, Zero-shot=true2022.05 | 63.3 | 89.3 | 95.7 | — | — | |
| R2D2Model Size=0.4B, Protocol=Zero-shot2024.01 | 63.3 | 89.3 | 95.7 | 80.5 | — | |
| CN-CLIPModel Size=1B, Protocol=Zero-shot2024.01 | 63 | 86.6 | 92.9 | 83 | — | |
| CN-CLIPModel Size=0.2B, Protocol=Zero-shot2024.01 | 62.2 | 86.6 | 94.9 | 79.7 | — | |
| FILIPBackbone=ViT-L, Zero-shot=true2022.02 | 56.9 | 82.4 | 90.9 | 75.2 | — | |
| FILIPBackbone=ViT-L, Zero-shot=true2022.05 | 56.9 | 82.4 | 90.9 | — | — | |
| WukongBackbone=ViT-L, Zero-shot=true2022.02 | 55.2 | 81 | 90.6 | 75.1 | — | |
| WukongSetting=Zero-shot2022.09 | 55.2 | 81 | 90.6 | — | 75.1 | |
| WukongBackbone=ViT-L, Zero-shot=true2022.05 | 55.2 | 81 | 90.6 | — | — | |
| WukongModel Size=0.4B, Protocol=Zero-shot2024.01 | 55.2 | 81 | 90.6 | 75.1 | — | |
| CLIPBackbone=ViT-B, Zero-shot=true2022.02 | 52.8 | 79.6 | 88.9 | 73 | — | |
| FILIP_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 52.7 | 81.3 | 88.3 | 76.6 | — | |
| TaisuModel Size=0.2B, Protocol=Zero-shot2024.01 | 52.5 | 81.5 | 91.4 | 75.9 | — | |
| CLIPBackbone=ViT-L, Zero-shot=true2022.02 | 51 | 80 | 89.7 | 72.1 | — | |
| CLIPBackbone=ViT-L, Zero-shot=true2022.05 | 51 | 80 | 89.7 | — | — | |
| CLIPBackbone=Swin-L, Zero-shot=true2022.02 | 50.5 | 79.2 | 88.2 | 71.7 | — | |
| FILIPBackbone=Swin-L, Zero-shot=true2022.02 | 48.6 | 77.3 | 88.3 | 72.1 | — | |
| WukongBackbone=ViT-B, Zero-shot=true2022.02 | 48.3 | 77.8 | 88.8 | 71.9 | — | |
| WukongBackbone=Swin-L, Zero-shot=true2022.02 | 47.3 | 78 | 88.3 | 70.8 | — | |
| FILIPBackbone=ViT-B, Zero-shot=true2022.02 | 37.8 | 66.4 | 77.9 | 61.8 | — | |
| BriVLBackbone=N/A, Zero-shot=true2022.02 | 17.1 | 41.7 | 57.5 | 37.4 | — | |
| BriVLModel Size=1B, Protocol=Zero-shot2024.01 | 17.1 | 41.7 | 57.5 | 37.4 | — | |
| EmbNEvaluation Protocol=Fine-tuned2022.02 | — | — | — | 73.2 | — | |
| LightningDOTEvaluation Protocol=Fine-tuned2022.02 | — | — | — | 88.4 | — | |
| M3PSetting=Zero-shot2022.09 | — | — | — | — | 32.3 | |
| M³PEvaluation Protocol=Fine-tuned2022.02 | — | — | — | 86.2 | — | |
| PARALLEL-EmbNEvaluation Protocol=Fine-tuned2022.02 | — | — | — | 76 | — | |
| S-LIWEEvaluation Protocol=Fine-tuned2022.02 | — | — | — | 73.6 | — | |
| UC2Evaluation Protocol=Fine-tuned2022.02 | — | — | — | 89.8 | — | |
| UNITEREvaluation Protocol=Fine-tuned2022.02 | — | — | — | 87.3 | — |