Composed Image Retrieval on FashionIQ Shirt
61.91Recall@10ReTrack
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ReTrack2026.04 | 61.91 | 81.26 | |
| IMAGINEModel Category=CVR Models2026.06 | 61.61 | 80.31 | |
| LIMN+2026.04 | 57.51 | 77.92 | |
| LIMN+Model Category=CIR Models2026.06 | 57.51 | 77.92 | |
| LIMN2026.04 | 56.08 | 77.09 | |
| LIMNModel Category=CIR Models2026.06 | 56.08 | 77.09 | |
| SPRC2026.04 | 55.64 | 73.89 | |
| SPRCModel Category=CIR Models2026.06 | 55.64 | 73.89 | |
| ENCODER2026.04 | 54.86 | 74.93 | |
| ENCODERModel Category=CIR Models2026.06 | 54.86 | 74.93 | |
| QuReModel Category=CIR Models2026.06 | 53.53 | 72.87 | |
| MEDIANModel Category=CIR Models2026.06 | 52.65 | 73.96 | |
| TG-CIR2026.04 | 52.6 | 72.52 | |
| TG-CIRModel Category=CIR Models2026.06 | 52.6 | 72.52 | |
| PAIRModel Category=CIR Models2026.06 | 52.6 | 73.8 | |
| CoVR-22026.04 | 51.23 | 70.64 | |
| CoVR-2Model Category=CVR Models2026.06 | 51.23 | 70.64 | |
| CoVR Enrich2026.04 | 49.61 | 68.88 | |
| CoVR_EnrichModel Category=CVR Models2026.06 | 49.61 | 68.88 | |
| CoVR2026.04 | 48.43 | 67.42 | |
| CoVRModel Category=CVR Models2026.06 | 48.43 | 67.42 | |
| LinCIRBackbone=ViT-G/14, Training-free=false2026.02 | 46.76 | 65.11 | |
| LinCIRArch=ViT-G/14, Train=true2026.05 | 46.76 | 65.11 | |
| QQMM-embed-v2Version=v22026.02 | 46.07 | 65.65 | |
| WISERBackbone=ViT-G/14, Training-free=true2026.02 | 45.98 | 61.58 | |
| RzenEmbed-v2-7BParameters=7B2026.02 | 45.63 | 64.97 | |
| OSCAR2026.02 | 44.5 | 67.52 | |
| SADN2026.04 | 43.67 | 66.05 | |
| SADNModel Category=CIR Models2026.06 | 43.67 | 66.05 | |
| SEIZEArch=ViT-G/14, Train=false2026.05 | 43.6 | 65.42 | |
| WISERBackbone=ViT-L/14, Training-free=true2026.02 | 43.13 | 57.85 | |
| IUDC2026.04 | 41.86 | 63.52 | |
| IUDCModel Category=CIR Models2026.06 | 41.86 | 63.52 | |
| WISERBackbone=ViT-B/32, Training-free=true2026.02 | 41.51 | 56.92 | |
| MRA-CIR2026.02 | 40.43 | 60.2 | |
| FiRE2026.02 | 39.84 | 60.06 | |
| STiTchArch=ViT-G/14, Train=false2026.05 | 39.48 | 56.59 | |
| X^R2026.02 | 38.91 | 56.82 | |
| OSrCIRBackbone=ViT-G/14, Training-free=true2026.02 | 38.65 | 54.71 | |
| OSrCIRArch=ViT-G/14, Train=false2026.05 | 38.65 | 54.71 | |
| CoTMRBackbone=ViT-G/14, Training-free=true2026.02 | 38.32 | 62.24 | |
| SSN2026.04 | 38.13 | 61.83 | |
| SEARLEArch=ViT-G/14, Train=true2026.05 | 36.46 | 55.35 | |
| AutoCIRBackbone=ViT-G/14, Training-free=true2026.02 | 36.36 | 55.84 | |
| LDREBackbone=ViT-G/14, Training-free=true2026.02 | 35.94 | 58.58 | |
| CoTMRBackbone=ViT-L/14, Training-free=true2026.02 | 35.43 | 54.91 | |
| AutoCIR2026.02 | 34 | 53.43 | |
| AutoCIRBackbone=ViT-L/14, Training-free=true2026.02 | 34 | 53.43 | |
| CIReVLBackbone=ViT-G/14, Training-free=true2026.02 | 33.71 | 51.42 | |
| CoTMRBackbone=ViT-B/32, Training-free=true2026.02 | 33.42 | 53.93 | |
| OSrCIRBackbone=ViT-L/14, Training-free=true2026.02 | 33.17 | 52.03 | |
| Pic2WordArch=ViT-G/14, Train=true2026.05 | 33.17 | 50.39 | |
| MagicLensBackbone=CLIP ViT-L/142026.07 | 32.7 | 53.8 | |
| AutoCIRBackbone=ViT-B/32, Training-free=true2026.02 | 32.43 | 51.67 | |
| HITBackbone=ViT-L/14, Training-free=false2026.02 | 32.4 | 51.2 | |
| MOABackbone=ViT-L/14, Training-free=false2026.02 | 31.9 | 50.7 | |
| Ops-MM-v1-7BParameters=7B2026.02 | 31.45 | 48.87 | |
| OSrCIRBackbone=ViT-B/32, Training-free=true2026.02 | 31.16 | 51.13 | |
| LDRE2026.02 | 31.04 | 51.22 | |
| LDREBackbone=ViT-L/14, Training-free=true2026.02 | 31.04 | 51.22 | |
| SEARLE-XL-OTIScale=XL-OTI2026.02 | 30.37 | 47.49 | |
| CIReVLArch=ViT-G/14, Train=false2026.05 | 29.85 | 51.07 | |
| Context-I2WBackbone=CLIP ViT-L/142026.07 | 29.7 | 48.6 | |
| CIReVL2026.02 | 29.49 | 47.4 | |
| CIReVLBackbone=ViT-L/14, Training-free=true2026.02 | 29.49 | 47.4 | |
| LinCIR2026.02 | 29.1 | 46.81 | |
| LinCIRBackbone=ViT-L/14, Training-free=false2026.02 | 29.1 | 46.81 | |
| LinCIRBackbone=CLIP ViT-L/142026.07 | 29.1 | 46.8 | |
| Qwen3-Embed-8BParameters=8B2026.02 | 28.7 | 43.13 | |
| CIReVLBackbone=ViT-B/32, Training-free=true2026.02 | 28.36 | 47.84 | |
| LDREBackbone=ViT-B/32, Training-free=true2026.02 | 27.38 | 46.27 | |
| MagicLensBackbone=CLIP ViT-B/322026.07 | 27.3 | 48.8 | |
| SEARLEBackbone=CLIP ViT-L/142026.07 | 26.9 | 45.6 | |
| SEARLE2026.02 | 26.89 | 45.58 | |
| SEARLEBackbone=ViT-L/14, Training-free=false2026.02 | 26.89 | 45.58 | |
| Qwen3-Embed-4BParameters=4B2026.02 | 26.69 | 40.68 | |
| Pic2Word2026.02 | 26.2 | 43.6 | |
| Pic2WordBackbone=CLIP ViT-L/142026.07 | 26.2 | 43.6 | |
| SEARLEBackbone=CLIP ViT-B/322026.07 | 24.8 | 41.1 | |
| SEARLEBackbone=ViT-B/32, Training-free=false2026.02 | 24.44 | 41.61 | |
| FlowCIRBackbone=CLIP ViT-L/142026.07 | 24.4 | 44.3 | |
| PALAVRABackbone=CLIP ViT-B/322026.07 | 21.5 | 37.1 | |
| PALAVRABackbone=ViT-B/32, Training-free=false2026.02 | 21.49 | 37.05 | |
| Qwen3-Embed-0.6BParameters=0.6B2026.02 | 21 | 34.69 | |
| bge-m32026.02 | 20.66 | 33.66 | |
| Text-onlyBackbone=CLIP ViT-B/322026.07 | 20.3 | 35.3 | |
| B3_Qwen2_7BBackbone=Qwen2, Parameters=7B2026.02 | 19.53 | 34.49 | |
| FlowCIRBackbone=CLIP ViT-B/322026.07 | 19.1 | 40.1 | |
| VLM2Vec2026.02 | 15.6 | 30.62 | |
| Text-onlyBackbone=CLIP ViT-L/142026.07 | 13.6 | 30 | |
| Image-onlyBackbone=CLIP ViT-B/322026.07 | 7.5 | 14 | |
| Image-onlyBackbone=CLIP ViT-L/142026.07 | 4.5 | 12.2 |