# Image-to-Text Retrieval on MS-COCO 5-fold 1K (test) Source: [Image-to-Text Retrieval on MS-COCO 5-fold 1K (test)]() Collection index: [llms.txt]() Data updated: 2026-07-03T06:49:30.879Z Evaluation of Image-to-Text Retrieval performance using the 5-fold 1K test subset of the MS-COCO dataset, measured using R@K metrics and RSUM. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Image-to-Text Retrieval | | Dataset (including any stated split) | MS-COCO 5-fold 1K (test) | | Metric | R@1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 35 of 35 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [R@1](): Higher is better; unit: PERCENT. - [R@5](): Higher is better; unit: PERCENT. - [R@10](): Higher is better; unit: PERCENT. - [RSUM](): Higher is better; unit: SCALAR. ## Recorded results | Method | R@1 | Result date | Report paper | | --- | --- | --- | --- | | SEPS | 90.9 | Not recorded | 690988853ef7ccd74e5e5c23 | | SEPS | 89.5 | Not recorded | 690988853ef7ccd74e5e5c23 | | SEPS | 89 | Not recorded | 690988853ef7ccd74e5e5c23 | | SEPS | 87.2 | Not recorded | 690988853ef7ccd74e5e5c23 | | HREM | 82.9 | Not recorded | 68fa73a71e209131cea5db16 | | CORA | 82.8 | Not recorded | 68fa73a71e209131cea5db16 | | CORA | 82.4 | Not recorded | 68fa73a71e209131cea5db16 | | CODER | 82.1 | Not recorded | 68fa73a71e209131cea5db16 | | CORA | 81.7 | Not recorded | 68fa73a71e209131cea5db16 | | CHAN | 81.4 | Not recorded | 68fa73a71e209131cea5db16 | | CORA | 81.2 | Not recorded | 68fa73a71e209131cea5db16 | | HREM | 81.2 | Not recorded | 68fa73a71e209131cea5db16 | | CORA | 80.9 | Not recorded | 68fa73a71e209131cea5db16 | | SDE | 80.6 | Not recorded | 68fa73a71e209131cea5db16 | | CORA | 80.5 | Not recorded | 68fa73a71e209131cea5db16 | | NAAF | 80.5 | Not recorded | 68fa73a71e209131cea5db16 | | MV-VSE | 80.4 | Not recorded | 68fa73a71e209131cea5db16 | | CHAN | 79.7 | Not recorded | 68fa73a71e209131cea5db16 | | VSE∞ | 79.7 | Not recorded | 68fa73a71e209131cea5db16 | | SGARF | 79.6 | Not recorded | 68fa73a71e209131cea5db16 | | SGARF | 79.6 | Not recorded | 68fa73a71e209131cea5db16 | | CODER | 78.9 | Not recorded | 68fa73a71e209131cea5db16 | | MV-VSE | 78.7 | Not recorded | 68fa73a71e209131cea5db16 | | MV-VSE | 78.7 | Not recorded | 68fa73a71e209131cea5db16 | | VSE∞ | 78.5 | Not recorded | 68fa73a71e209131cea5db16 | | VSE∞ | 78.5 | Not recorded | 68fa73a71e209131cea5db16 | | GraDual | 77 | Not recorded | 68fa73a71e209131cea5db16 | | VSRN | 76.2 | Not recorded | 68fa73a71e209131cea5db16 | | VSRN | 76.2 | Not recorded | 68fa73a71e209131cea5db16 | | CAAN | 75.5 | Not recorded | 68fa73a71e209131cea5db16 | | CAAN | 75.5 | Not recorded | 68fa73a71e209131cea5db16 | | SGM | 73.4 | Not recorded | 68fa73a71e209131cea5db16 | | SGM | 73.4 | Not recorded | 68fa73a71e209131cea5db16 | | SCAN | 72.7 | Not recorded | 68fa73a71e209131cea5db16 | | SCAN | 72.7 | Not recorded | 68fa73a71e209131cea5db16 | ### SEPS R@1: 90.9. Result ID: 6a475b5c60b88bb339922c65. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Encoder":"ViT-Base-384","Textual Encoder":"BERT-base","Image Resolution":"384x384","Patch Size":"24x24","Fine-Grained Alignment (FG)":"true"}. This record does not assert independent reproduction. ### SEPS R@1: 89.5. Result ID: 6a475b5c60b88bb339922c67. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Encoder":"Swin-Base-384","Textual Encoder":"BERT-base","Image Resolution":"384x384","Patch Size":"12x12","Fine-Grained Alignment (FG)":"true"}. This record does not assert independent reproduction. ### SEPS R@1: 89. Result ID: 6a475b5c60b88bb339922c64. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Encoder":"ViT-Base-224","Textual Encoder":"BERT-base","Image Resolution":"224x224","Patch Size":"14x14","Fine-Grained Alignment (FG)":"true"}. This record does not assert independent reproduction. ### SEPS R@1: 87.2. Result ID: 6a475b5c60b88bb339922c66. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Encoder":"Swin-Base-224","Textual Encoder":"BERT-base","Image Resolution":"224x224","Patch Size":"7x7","Fine-Grained Alignment (FG)":"true"}. This record does not assert independent reproduction. ### HREM R@1: 82.9. Result ID: 69a1624035dfcabb00e6b013. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"false","Ensemble":"true","Venue":"CVPR’23"}. This record does not assert independent reproduction. ### CORA R@1: 82.8. Result ID: 69a1624035dfcabb00e6b015. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"false","Ensemble":"true"}. This record does not assert independent reproduction. ### CORA R@1: 82.4. Result ID: 69a1624035dfcabb00e6b014. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"false","Ensemble":"false"}. This record does not assert independent reproduction. ### CODER R@1: 82.1. Result ID: 69a1624035dfcabb00e6b010. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"true","Ensemble":"false","Venue":"ECCV’22"}. This record does not assert independent reproduction. ### CORA R@1: 81.7. Result ID: 69a1624035dfcabb00e6b00e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"true"}. This record does not assert independent reproduction. ### CHAN R@1: 81.4. Result ID: 69a1624035dfcabb00e6b012. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"true","Ensemble":"false","Venue":"CVPR’23"}. This record does not assert independent reproduction. ### CORA R@1: 81.2. Result ID: 69a1624035dfcabb00e6afac. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Cross-Attention":"false","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"true"}. This record does not assert independent reproduction. ### HREM R@1: 81.2. Result ID: 69a1624035dfcabb00e6b00c. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"true","Venue":"CVPR’23"}. This record does not assert independent reproduction. ### CORA R@1: 80.9. Result ID: 69a1624035dfcabb00e6b00d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"false"}. This record does not assert independent reproduction. ### SDE R@1: 80.6. Result ID: 69a1624035dfcabb00e6b00b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"true","Venue":"CVPR’23"}. This record does not assert independent reproduction. ### CORA R@1: 80.5. Result ID: 69a1624035dfcabb00e6afab. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Cross-Attention":"false","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"false"}. This record does not assert independent reproduction. ### NAAF R@1: 80.5. Result ID: 69a1624035dfcabb00e6b00a. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"true","Venue":"CVPR’23"}. This record does not assert independent reproduction. ### MV-VSE R@1: 80.4. Result ID: 69a1624035dfcabb00e6b011. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"false","Ensemble":"true","Venue":"IJCAI’22"}. This record does not assert independent reproduction. ### CHAN R@1: 79.7. Result ID: 69a1624035dfcabb00e6b009. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"false","Venue":"CVPR’23"}. This record does not assert independent reproduction. ### VSE∞ R@1: 79.7. Result ID: 69a1624035dfcabb00e6b00f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"BERT","Cross Attention":"false","Ensemble":"false","Venue":"CVPR’21"}. This record does not assert independent reproduction. ### SGARF R@1: 79.6. Result ID: 69a1624035dfcabb00e6afa9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Venue":"AAAI’21","Cross-Attention":"true","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"true"}. This record does not assert independent reproduction. ### SGARF R@1: 79.6. Result ID: 69a1624035dfcabb00e6b005. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"true","Venue":"AAAI’23"}. This record does not assert independent reproduction. ### CODER R@1: 78.9. Result ID: 69a1624035dfcabb00e6b006. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"false","Venue":"ECCV’22"}. This record does not assert independent reproduction. ### MV-VSE R@1: 78.7. Result ID: 69a1624035dfcabb00e6afaa. [Reporting paper]() Recorded evaluation configuration: {"Venue":"IJCAI’22","Cross-Attention":"false","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"true"}. This record does not assert independent reproduction. ### MV-VSE R@1: 78.7. Result ID: 69a1624035dfcabb00e6b007. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"true","Venue":"IJCAI’22"}. This record does not assert independent reproduction. ### VSE∞ R@1: 78.5. Result ID: 69a1624035dfcabb00e6afa8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Venue":"CVPR’21","Cross-Attention":"false","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"false"}. This record does not assert independent reproduction. ### VSE∞ R@1: 78.5. Result ID: 69a1624035dfcabb00e6b004. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"false","Venue":"CVPR’21"}. This record does not assert independent reproduction. ### GraDual R@1: 77. Result ID: 69a1624035dfcabb00e6b008. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"true","Venue":"WACV’23"}. This record does not assert independent reproduction. ### VSRN R@1: 76.2. Result ID: 69a1624035dfcabb00e6afa5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Venue":"ICCV’19","Cross-Attention":"false","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"false"}. This record does not assert independent reproduction. ### VSRN R@1: 76.2. Result ID: 69a1624035dfcabb00e6b001. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"false","Ensemble":"false","Venue":"ICCV’19"}. This record does not assert independent reproduction. ### CAAN R@1: 75.5. Result ID: 69a1624035dfcabb00e6afa7. [Reporting paper]() Recorded evaluation configuration: {"Venue":"CVPR’20","Cross-Attention":"true","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"false"}. This record does not assert independent reproduction. ### CAAN R@1: 75.5. Result ID: 69a1624035dfcabb00e6b003. [Reporting paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"false","Venue":"CVPR’20"}. This record does not assert independent reproduction. ### SGM R@1: 73.4. Result ID: 69a1624035dfcabb00e6afa6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Venue":"WACV’20","Cross-Attention":"true","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"false"}. This record does not assert independent reproduction. ### SGM R@1: 73.4. Result ID: 69a1624035dfcabb00e6b002. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"false","Venue":"WACV’20"}. This record does not assert independent reproduction. ### SCAN R@1: 72.7. Result ID: 69a1624035dfcabb00e6afa4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Venue":"ECCV’18","Cross-Attention":"true","Backbone":"Faster R-CNN + Bi-GRU","Ensemble":"true"}. This record does not assert independent reproduction. ### SCAN R@1: 72.7. Result ID: 69a1624035dfcabb00e6b000. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Visual Backbone":"Faster R-CNN","Semantic Encoder":"Bi-GRU","Cross Attention":"true","Ensemble":"true","Venue":"ECCV’18"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.