Composed Image Retrieval on FashionIQ (Dress)
52.91Recall@10ReTrack
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ReTrack2026.04 | 52.91 | 77.54 | |
| IMAGINEModel Category=CVR Models2026.06 | 52.39 | 77.04 | |
| LIMN+2026.04 | 52.11 | 75.21 | |
| LIMN+Model Category=CIR Models2026.06 | 52.11 | 75.21 | |
| ENCODER2026.04 | 51.51 | 76.95 | |
| ENCODERModel Category=CIR Models2026.06 | 51.51 | 76.95 | |
| LIMN2026.04 | 50.72 | 74.52 | |
| LIMNModel Category=CIR Models2026.06 | 50.72 | 74.52 | |
| SPRC2026.04 | 49.18 | 72.43 | |
| SPRCModel Category=CIR Models2026.06 | 49.18 | 72.43 | |
| MEDIANModel Category=CIR Models2026.06 | 46.9 | 70.3 | |
| QuReModel Category=CIR Models2026.06 | 46.8 | 69.81 | |
| PAIRModel Category=CIR Models2026.06 | 46.78 | 70.93 | |
| CoVR-22026.04 | 46.53 | 69.6 | |
| CoVR-2Model Category=CVR Models2026.06 | 46.53 | 69.6 | |
| CoVR Enrich2026.04 | 46.12 | 69.52 | |
| CoVR_EnrichModel Category=CVR Models2026.06 | 46.12 | 69.52 | |
| TG-CIR2026.04 | 45.22 | 69.66 | |
| TG-CIRModel Category=CIR Models2026.06 | 45.22 | 69.66 | |
| CoVR2026.04 | 44.55 | 69.03 | |
| CoVRModel Category=CVR Models2026.06 | 44.55 | 69.03 | |
| SADN2026.04 | 40.01 | 65.1 | |
| SADNModel Category=CIR Models2026.06 | 40.01 | 65.1 | |
| SEIZEArch=ViT-G/14, Train=false2026.05 | 39.61 | 61.02 | |
| OSCAR2026.02 | 38.47 | 65.15 | |
| LinCIRArch=ViT-G/14, Train=true2026.05 | 38.08 | 60.88 | |
| RzenEmbed-v2-7BParameters=7B2026.02 | 37.38 | 61.97 | |
| QQMM-embed-v2Version=v22026.02 | 36.44 | 60.09 | |
| IUDC2026.04 | 35.22 | 61.9 | |
| IUDCModel Category=CIR Models2026.06 | 35.22 | 61.9 | |
| STiTchArch=ViT-G/14, Train=false2026.05 | 35.04 | 56.74 | |
| SSN2026.04 | 34.36 | 60.78 | |
| OSrCIRArch=ViT-G/14, Train=false2026.05 | 33.02 | 54.78 | |
| MRA-CIR2026.02 | 31.87 | 54.23 | |
| FlowCIRBackbone=CLIP ViT-L/142026.07 | 31.6 | 48.5 | |
| FiRE2026.02 | 29.6 | 50.87 | |
| X^R2026.02 | 28.71 | 52.5 | |
| SEARLEArch=ViT-G/14, Train=true2026.05 | 28.16 | 50.32 | |
| CIReVLArch=ViT-G/14, Train=false2026.05 | 27.07 | 49.53 | |
| MagicLensBackbone=CLIP ViT-L/142026.07 | 25.5 | 46.1 | |
| Pic2WordArch=ViT-G/14, Train=true2026.05 | 25.43 | 47.65 | |
| AutoCIR2026.02 | 24.94 | 45.81 | |
| FlowCIRBackbone=CLIP ViT-B/322026.07 | 24.8 | 41.8 | |
| CIReVL2026.02 | 24.79 | 44.76 | |
| Context-I2WBackbone=CLIP ViT-L/142026.07 | 23.1 | 45.3 | |
| LDRE2026.02 | 22.93 | 46.76 | |
| SEARLE-XL-OTIScale=XL-OTI2026.02 | 21.57 | 44.47 | |
| MagicLensBackbone=CLIP ViT-B/322026.07 | 21.5 | 41.3 | |
| LinCIR2026.02 | 20.92 | 42.44 | |
| LinCIRBackbone=CLIP ViT-L/142026.07 | 20.9 | 42.4 | |
| SEARLEBackbone=CLIP ViT-L/142026.07 | 20.5 | 43.1 | |
| SEARLE2026.02 | 20.48 | 43.13 | |
| Pic2Word2026.02 | 20.2 | 40.2 | |
| Pic2WordBackbone=CLIP ViT-L/142026.07 | 20 | 40.2 | |
| Ops-MM-v1-7BParameters=7B2026.02 | 19.39 | 38.13 | |
| Text-onlyBackbone=CLIP ViT-L/142026.07 | 18.3 | 30.1 | |
| SEARLEBackbone=CLIP ViT-B/322026.07 | 18.2 | 38.6 | |
| PALAVRABackbone=CLIP ViT-B/322026.07 | 17.3 | 35.9 | |
| Text-onlyBackbone=CLIP ViT-B/322026.07 | 13.6 | 31.8 | |
| Qwen3-Embed-4BParameters=4B2026.02 | 12.69 | 27.02 | |
| Qwen3-Embed-8BParameters=8B2026.02 | 12.64 | 29.2 | |
| bge-m32026.02 | 10.81 | 23.75 | |
| Image-onlyBackbone=CLIP ViT-L/142026.07 | 10.7 | 19.9 | |
| Qwen3-Embed-0.6BParameters=0.6B2026.02 | 9.32 | 19.83 | |
| B3_Qwen2_7BBackbone=Qwen2, Parameters=7B2026.02 | 8.53 | 22.31 | |
| VLM2Vec2026.02 | 4.76 | 14.77 | |
| Image-onlyBackbone=CLIP ViT-B/322026.07 | 3.9 | 10.8 |