Image Retrieval on Flickr30K 1K (test)
91.1R@1X-FMbase
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| X-FMbaseParameters=327M, Evaluation Protocol=Fine-Tuned, Training Data=More Data2023.01 | 91.1 | 98.5 | 99.4 | — | |
| X2-VLMbaseParameters=255M, Evaluation Protocol=Fine-Tuned, Training Data=More Data2023.01 | 90.4 | 98.2 | 99.3 | — | |
| CoCaParameters=2.1B, Evaluation Protocol=Fine-Tuned, Training Data=Super-Large2023.01 | 90.3 | 98.7 | 99.5 | — | |
| X-FMbaseParameters=327M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 90 | 98.6 | 99.3 | — | |
| X2-VLMbaseParameters=255M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 88.6 | 97.9 | 98.9 | — | |
| mPLUG-2Base#PT Data=17M2023.02 | 88.2 | 97.8 | 99 | — | |
| mPLUG-2#PT Data=17M2023.02 | 88.1 | 97.6 | 99.1 | — | |
| Florence#PT Data=0.9B2023.02 | 87.9 | 98.1 | — | — | |
| Florence# Img-Text Pairs=900M, Fine-tuned=true2022.09 | 87.9 | 98.1 | — | — | |
| OmniVL# Img-Text Pairs=14M*, Fine-tuned=true2022.09 | 87.9 | 97.8 | 99.1 | — | |
| BLIP#PT Data=129M2023.02 | 87.6 | 97.7 | 99 | — | |
| BLIP2023.05 | 87.6 | 97.7 | 99 | — | |
| BLIP# Img-Text Pairs=14M, Fine-tuned=true2022.09 | 87.2 | 97.5 | 98.8 | — | |
| X-VLMParameters=216M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 86.1 | 97.4 | 98.7 | — | |
| RaSa2023.05 | 85.9 | 97.54 | 98.72 | — | |
| ALBEF#PT Data=14M2023.02 | 85.6 | 97.5 | 98.9 | — | |
| ALBEF# Img-Text Pairs=14M, Fine-tuned=true2022.09 | 85.6 | 97.5 | 98.9 | — | |
| ALBEFconfiguration=backbone2023.05 | 85.44 | 97.34 | 98.7 | — | |
| METER + EPICData=4M, EPIC=true, Protocol=Fine-tuned2022.11 | 85.1 | 97.4 | 98.7 | — | |
| ALIGN#PT Data=1.8B2023.02 | 84.9 | 97.4 | 98.6 | — | |
| ALIGN2023.05 | 84.9 | 97.4 | 98.6 | — | |
| ALIGN# Img-Text Pairs=1.8B, Fine-tuned=true2022.09 | 84.9 | 97.4 | 98.6 | — | |
| VLMo#PT Data=4M2023.02 | 84.5 | 97.3 | 98.6 | — | |
| X-FMbaseParameters=327M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.01 | 84.1 | 96.5 | 98.1 | — | |
| VL-BEITParameters=175M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 83.9 | — | — | — | |
| METERData=4M, EPIC=false, Protocol=Fine-tuned2022.11 | 83.8 | 97.3 | 98.6 | — | |
| OmniVLParameters=288M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 83.4 | 97 | 98.6 | — | |
| OmniVL# Img-Text Pairs=4M*, Fine-tuned=true2022.09 | 83.4 | 97 | 98.6 | — | |
| ALBEFParameters=210M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 82.8 | 96.7 | 98.4 | — | |
| METER# Img-Text Pairs=404M, Fine-tuned=true2022.09 | 82.2 | 96.3 | 98.4 | — | |
| CoCaParameters=2.1B, Evaluation Protocol=Zero-Shot, Training Data=Super-Large2023.01 | 80.4 | 95.7 | 97.7 | — | |
| VLMbaseParameters=175M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.01 | 79.3 | 95.7 | 97.8 | — | |
| VLMO# Img-Text Pairs=4M, Fine-tuned=true2022.09 | 79.3 | 95.7 | 97.8 | — | |
| X2-VLMbaseParameters=255M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.01 | 79.2 | 96.4 | 98 | — | |
| X2-VLMbaseParameters=255M, Evaluation Protocol=Zero-Shot, Training Data=Standard2023.01 | 79.1 | 95.2 | 97.3 | — | |
| METER + EPICData=4M, EPIC=true, Protocol=Zero-shot2022.11 | 79 | 94.5 | 97.1 | — | |
| METERData=4M, EPIC=false, Protocol=Zero-shot2022.11 | 78.8 | 94.5 | 96.9 | — | |
| UNIMO2023.05 | 78.04 | 94.24 | 97.12 | — | |
| UNIMO#PT Data=4M2023.02 | 78 | 94.2 | 97.1 | — | |
| UNIMO# Img-Text Pairs=4M, Fine-tuned=true2022.09 | 78 | 94.2 | 97.1 | — | |
| X-FMbaseParameters=327M, Evaluation Protocol=Zero-Shot, Training Data=Standard2023.01 | 77.3 | 95.3 | 97.6 | — | |
| FlorenceParameters=893M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.01 | 76.7 | 93.6 | — | — | |
| ALIGNParameters=490M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.01 | 75.7 | 93.8 | 96.8 | — | |
| UNITER#PT Data=4M2023.02 | 75.6 | 94.1 | 96.8 | — | |
| COOKIE2023.05 | 75.6 | 94.6 | 97.2 | — | |
| UNITER# Img-Text Pairs=4M, Fine-tuned=true2022.09 | 75.6 | 94.1 | 96.8 | — | |
| UNITER2023.05 | 75.56 | 94.08 | 96.76 | — | |
| E2E-VLP#PT Data=4M2023.02 | 73.6 | 92.4 | 96 | — | |
| FLAVA# Img-Text Pairs=70M, Fine-tuned=true2022.09 | 73.2 | 92.7 | 95.5 | — | |
| X-VLMParameters=216M, Evaluation Protocol=Zero-Shot, Training Data=Standard2023.01 | 71.9 | 93.3 | 96.4 | — | |
| CLIPParameters=490M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.01 | 68.7 | 90.6 | 95.2 | — | |
| RCARWarm-up=true, Model Variant=All2023.03 | 62.6 | 85.8 | 91.1 | — | |
| RCARWarm-up=true, Model Variant=T2I2023.03 | 60.9 | 84.4 | 90.1 | — | |
| PCSR (AAAI'24)Noise=0.22026.06 | 60.9 | 83.7 | 89.4 | 505.7 | |
| GPOWarm-up=true, Ensemble=true2023.03 | 60.8 | 86.3 | 92.3 | — | |
| SPS (IJCAI'24)Noise=0.22026.06 | 60.5 | 84.3 | 89.8 | 507.1 | |
| IN2RNoise=0.22026.06 | 60.5 | 86.2 | 91.7 | 511.6 | |
| BiCro (CVPR'23)Noise=0.22026.06 | 60.4 | 84.4 | 89.9 | 504.7 | |
| GSC (CVPR'24)Noise=0.22026.06 | 60.1 | 84.5 | 90.5 | 505.8 | |
| L2RM (CVPR'24)Noise=0.22026.06 | 59.8 | 83.6 | 89.5 | 503.8 | |
| RCARModel Variant=All2023.03 | 59.5 | 84 | 89.5 | — | |
| ESC (CVPR'24)Noise=0.22026.06 | 59.1 | 83.8 | 89.1 | 503.3 | |
| RCARWarm-up=true, Model Variant=I2T2023.03 | 58.8 | 83.9 | 89.3 | — | |
| CREAM (TIP'24)Noise=0.22026.06 | 58.7 | 84.1 | 89.8 | 502.3 | |
| IN2RNoise=0.42026.06 | 58.7 | 84.7 | 90.7 | 504.1 | |
| SGRAFBackbone=ResNet-101 BUTD [5]2023.03 | 58.5 | 83 | 88.8 | — | |
| GPOWarm-up=true2023.03 | 58.3 | 84.6 | 90.8 | — | |
| GSC (CVPR'24)Noise=0.42026.06 | 57.5 | 82.7 | 88.9 | 497.3 | |
| GSMNBackbone=ResNet-101 BUTD [5]2023.03 | 57.4 | 82.3 | 89 | — | |
| PCSR (AAAI'24)Noise=0.42026.06 | 57.4 | 82.4 | 88.9 | 497.1 | |
| SPS (IJCAI'24)Noise=0.42026.06 | 57.3 | 83.5 | 89.6 | 498.9 | |
| RCARModel Variant=T2I2023.03 | 57.2 | 82.8 | 88.5 | — | |
| CREAM (TIP'24)Noise=0.42026.06 | 57 | 82.6 | 88.7 | 495 | |
| NCR (NeurIPS'21)Noise=0.22026.06 | 56.9 | 82.4 | 88.5 | 491.1 | |
| IN2RNoise=0.62026.06 | 56.5 | 82.9 | 89.5 | 495.2 | |
| L2RM (CVPR'24)Noise=0.42026.06 | 56.3 | 81 | 87.3 | 490.5 | |
| NCR (NeurIPS'21)Noise=0.42026.06 | 56.2 | 80.6 | 87.4 | 486.8 | |
| ESC (CVPR'24)Noise=0.42026.06 | 56 | 80.8 | 87.2 | 489.6 | |
| BiCro (CVPR'23)Noise=0.42026.06 | 55.5 | 81 | 87.4 | 487.5 | |
| SHANBackbone=ResNet-101 BUTD [5]2023.03 | 55.3 | 81.3 | 88.4 | — | |
| MMCABackbone=ResNet-101 BUTD [5]2023.03 | 54.8 | 81.4 | 87.8 | — | |
| WCGLBackbone=ResNet-101 BUTD [5]2023.03 | 54.8 | 80.6 | 87.5 | — | |
| PCSR (AAAI'24)Noise=0.62026.06 | 54.8 | 79.7 | 86.5 | 480.5 | |
| VSRNBackbone=ResNet-101 BUTD [5]2023.03 | 54.7 | 81.8 | 88.2 | — | |
| RCARModel Variant=I2T2023.03 | 54.6 | 80.5 | 87 | — | |
| IMRAMBackbone=ResNet-101 BUTD [5]2023.03 | 53.9 | 79.4 | 87.2 | — | |
| SPS (IJCAI'24)Noise=0.62026.06 | 53.7 | 80.2 | 87.7 | 484.1 | |
| SCANWarm-up=true2023.03 | 53.6 | 81.2 | 88.6 | — | |
| GSC (CVPR'24)Noise=0.62026.06 | 53.6 | 79.8 | 86.8 | 478 | |
| CREAM (TIP'24)Noise=0.62026.06 | 53.3 | 79.2 | 87 | 477.4 | |
| ESC (CVPR'24)Noise=0.62026.06 | 53 | 78.6 | 85.3 | 475 | |
| CAANBackbone=ResNet-101 BUTD [5]2023.03 | 52.8 | 79 | 87.9 | — | |
| NCR (NeurIPS'21)Noise=0.62026.06 | 52 | 77.6 | 84.9 | 468.6 | |
| L2RM (CVPR'24)Noise=0.62026.06 | 51.3 | 76.4 | 83.7 | 467.6 | |
| BiCro (CVPR'23)Noise=0.62026.06 | 51.2 | 77.6 | 84.7 | 466.3 | |
| SCANBackbone=ResNet-101 BUTD [5]2023.03 | 48.6 | 77.7 | 85.2 | — | |
| IN2RNoise=0.82026.06 | 48.2 | 76.4 | 84.7 | 458.8 | |
| PCSR (AAAI'24)Noise=0.82026.06 | 44.6 | 70.9 | 78.6 | 437.5 | |
| CREAM (TIP'24)Noise=0.82026.06 | 40.8 | 67.1 | 76.3 | 416.3 | |
| L2RM (CVPR'24)Noise=0.82026.06 | 39.4 | 65.4 | 74.9 | 404 |