List your company

Composed Image Retrieval on Fashion-IQ

82.01Average Recall@50MCoT-MVS
47.471656.438365.40574.3717Jun 13, 2024Oct 13, 2024Feb 13, 2025Jun 15, 2025Oct 16, 2025Feb 15, 2026Jun 18, 2026

Evaluation results

MethodLinks
MCoT-MVSMethod category=VLP Model-Based Methods2026.0382.0158.4578.9263.2481.1568.0285.9763.24—————————
DQU-CIRMethod category=VLP Model-Based Methods, Venue=SIGIR'242026.0381.5657.6378.5662.1480.3866.1585.7361.97—————————
HINT2026.0377.9351.873.1360.2778.4863.3782.1958.48—————————
ENCODERMethod category=VLP Model-Based Methods, Venue=AAAI'252026.0377.5951.5176.9554.8674.9362.0180.8856.13—————————
MAPNet2026.0676.251.1774.1256.3775.1759.5679.355.7—————————
CIRLVLMMethod category=VLP Model-Based Methods, Venue=AAAI'252026.0376.1450.4273.5758.5975.8659.6178.9956.21—————————
CLIP-4CIRCaptioning=LLaVA-enhanced2026.0675.6749.6271.8357.0275.6659.3179.555.32—————————
SPRC2026.0675.3547.872.755.8474.3758.8978.9954.17—————————
HABITNoise=0%2026.0475.249.9972.3856.6274.6859.5178.5355.38——65.29——————
TMENoise=0%2026.0475.0249.7371.6956.4374.4459.3178.9455.15——65.09——————
SPRCMethod category=VLP Model-Based Methods, Venue=ICLR'242026.0374.9749.1872.4355.6473.8959.3578.5854.72—————————
SPRC2026.0374.9749.1872.4355.6473.8959.3578.5854.72—————————
SPRCNoise=0%2026.0474.9749.1872.4355.6473.8959.3578.5854.92——64.85——————
CCINMethod category=VLP Model-Based Methods, Venue=CVPR'252026.0374.7649.3872.5855.9374.1457.9377.5654.41—————————
RCLNoise=0%2026.0474.6648.7972.6855.8973.956.9177.4153.86——64.26——————
PAIR2026.0374.5146.7870.9352.673.858.9178.8152.76—————————
MEDIAN2026.0374.346.970.352.6573.9657.6278.6352.39—————————
HABITNoise=20%2026.0474.2849.6371.3455.6773.1958.1478.3254.48——64.38——————
RDENoise=0%2026.0474.2247.8471.8954.3773.5556.9177.2153.04——63.63——————
TMENoise=20%2026.0473.9149.0370.3555.8473.1657.2278.2354.03——63.97——————
QuRe2026.0373.4846.869.8153.5372.8757.4777.7752.6—————————
TGCIRMethod category=VLP Model-Based Methods, Venue=ACM MM'232026.0373.0945.2269.6652.672.5256.1477.151.32—————————
TG-CIR2026.0373.0945.2269.6652.672.5256.1477.151.32—————————
HABITNoise=50%2026.0473.0947.3369.7153.7272.5556.517752.52——62.8——————
FashionERNMethod category=VLP Model-Based Methods, Venue=AAAI'242026.0372.9550.3271.2950.1570.3656.477.2152.29—————————
RCLNoise=20%2026.0472.6747.0570.6553.1471.7455.2875.6251.82——62.25——————
TMENoise=50%2026.0472.2546.2668.2753.0971.8855.0776.5951.47——61.86——————
RDENoise=20%2026.0470.6444.6268.9150.7469.0952.1273.3849.16——59.81——————
CASEMethod category=VLP Model-Based Methods, Venue=AAAI'242026.0370.6147.4469.3648.4870.2350.1872.2448.7—————————
CoVRBLIPMethod category=VLP Model-Based Methods, Venue=AAAI'242026.0370.2544.5569.0348.4367.4252.674.3148.53—————————
RCLNoise=50%2026.0469.8243.6866.4450.7469.1952.6373.8449.01——59.42——————
HABITNoise=80%2026.0469.5342.0465.250.1269.7752.9273.6148.36——58.94——————
CaLaMethod category=VLP Model-Based Methods, Venue=SIGIR'242026.0369.2242.3866.0846.7668.1650.9373.4246.69—————————
CALA2026.0369.2242.3866.0846.7668.1650.9373.4246.69—————————
CALANoise=0%2026.0469.2242.3866.0846.7668.1650.9373.4246.69——57.96——————
COPE2026.0368.639.8566.9845.0366.8148.6172.0144.5—————————
TMENoise=80%2026.0468.641.4564.3547.368.251.2573.2346.67——57.63——————
MMF2026.0367.5240.4565.2944.8566.2949.2670.9844.85—————————
SADNMethod category=VLP Model-Based Methods, Venue=ACM MM'242026.0367.3640.0165.143.6766.0548.0470.9343.91—————————
LinCIR + RTDBackbone=CLIP ViT-G/14, Integration with RTD=true2024.0667.3——————46.2—————————
RDENoise=50%2026.0467.2441.364.7547.0666.3450.1370.6346.16——56.7——————
SPRCNoise=20%2026.0466.3639.8162.2248.5866.2950.4870.5846.29——56.33——————
SEIZEArch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenCLIP2026.0565.8539.6161.0243.665.4245.9471.1243.05—————————
LinCIRArch=ViT-G/14, Train=true, Pre-trained Weights Source=OpenCLIP2026.0565.6938.0860.8846.7665.1150.4871.0945.11—————————
LinCIRBackbone=CLIP ViT-G/14, Integration with RTD=false2024.0665.5——————44.5—————————
IUDC2026.0364.8835.2261.941.8663.5242.1969.2339.76—————————
RCLNoise=80%2026.0464.4338.8260.5445.4464.3847.4268.3843.89——54.16——————
SSNMethod category=VLP Model-Based Methods, Venue=AAAI'242026.0363.8934.3660.7838.1361.8344.2669.0538.92—————————
SSN2026.0363.8934.3660.7838.1361.8344.2669.0538.92—————————
SSNNoise=0%2026.0463.8934.3660.7838.1361.8344.2669.0538.92——51.4——————
RDENoise=80%2026.0462.7537.6359.6443.6262.1246.166.542.45——52.6——————
Baseline (ViT-G)Pretraining Data=OACIRR, Pretraining Scale=127 K, Zero-shot=true2026.0461.81——————39.8—————————
CLIP4CIRMethod category=VLP Model-Based Methods, Venue=TOMM'232026.0361.7433.8159.439.9960.4541.4165.3738.4—————————
SPRCNoise=50%2026.0461.1935.9457.1642.2561.6344.9864.7641.06——51.12——————
CoAlign (ViT-G)Pretraining Data=CIRHS, Pretraining Scale=535 K, Zero-shot=true2026.0460.08——————39.22—————————
Baseline (ViT-L)Pretraining Data=OACIRR, Pretraining Scale=127 K, Zero-shot=true2026.0459.57——————37.82—————————
STiTch(Ours)Arch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenCLIP2026.0559.4335.0456.7439.4856.5942.8664.9539.12—————————
CompoDiff2026.0658.5740.6557.1436.8757.3943.9361.1740.48—————————
MT-CIRBackbone=CLIP ViT-L/142024.0657.4——————35.4—————————
PLIBackbone=CLIP-L, # Params=428M, # Triplets=695K, Zero-shot=true2025.1157.4——————35.4—————————
OSrCIR†Arch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenCLIP (rerun)2026.0557.1530.5953.537.3956.9239.7261.0435.79—————————
OSrCIRArch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenCLIP2026.0557.1133.0254.7838.6554.7141.0461.8337.57—————————
PVLF#Triplets=32K, Zero-shot=true, Backbone=BLIP-base2025.1156.9——————35.7—————————
OSrCIR*Arch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenAI (rerun)2026.0556.6730.6953.2536.5655.4540.1361.335.79—————————
CoLLMBackbone=BLIP-B, # Triplets=3.4M, Zero-shot=true2025.1156——————34.6—————————
CoLLM#Triplets=3.4M, Zero-shot=true, Backbone=BLIP-base2025.1156——————34.6—————————
SEARLEArch=ViT-G/14, Train=true, Pre-trained Weights Source=OpenCLIP2026.0555.7128.1650.3236.4655.3539.8361.4534.81—————————
LDREBackbone=CLIP ViT-G/142024.0655.4——————32.5—————————
CALANoise=20%2026.0455.2829.0551.3635.2856.2336.0558.2433.46——44.37——————
TransAggBackbone=BLIP-B, # Params=235M, # Triplets=32K, Zero-shot=true2025.1155.1——————34.4—————————
TransAgg#Triplets=32K, Zero-shot=true, Backbone=BLIP-base2025.1155.1——————34.4—————————
OSrCIR†Arch=ViT-B/32, Train=false, Pre-trained Weights Source=OpenCLIP (rerun)2026.0554.0729.7551.9132.8352.0636.3158.2432.96—————————
SPRCNoise=80%2026.0454.0328.4150.7736.2154.3735.959.0633.51——43.77——————
OSrCIRArch=ViT-B/32, Train=false, Pre-trained Weights Source=OpenAI2026.0553.429.3550.3731.1651.1336.5158.7132.34—————————
MagicLensBackbone=CLIP ViT-L/142024.0652.5——————30.7—————————
MagicLensBackbone=CLIP-L, # Params=465M, # Triplets=36.5M, Zero-shot=true2025.1152.5——————30.7—————————
CIReVLBackbone=CLIP ViT-G/142024.0652.4——————32.2—————————
CIReVLArch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenCLIP2026.0552.3627.0749.5329.8551.0735.856.1432.19—————————
LinCIR + RTDBackbone=FSC-CLIP ViT-L/14, Integration with RTD=true2024.0652.3——————31.2—————————
Pic2WordArch=ViT-G/14, Train=true, Pre-trained Weights Source=OpenCLIP2026.0551.8925.4347.6533.1750.3935.2457.6231.28—————————
CompoDiffPretraining Data=ST18M + LAION-2B, Pretraining Scale=18,000 K, Zero-shot=true2026.0451.71——————39.02—————————
FTI4CIR + RTDBackbone=CLIP ViT-L/14, Integration with RTD=true2024.0651.7——————30.6—————————
CompodiffBackbone=CLIP ViT-G/142024.0651.7——————39—————————
TransAgg + UNIONBackbone=BLIP-B, # Params=235M, # Triplets=5K, Zero-shot=true2025.1151.5——————31.9—————————
LinCIR + RTDBackbone=CLIP ViT-L/14, Integration with RTD=true2024.0651.1——————30.2—————————
FTI4CIRBackbone=CLIP ViT-L/14, Integration with RTD=false2024.0650.9——————29.4—————————
SEARLE + RTDBackbone=CLIP ViT-L/14, Integration with RTD=true2024.0650.7——————29.3—————————
LDREBackbone=CLIP ViT-L/142024.0650.5——————28.5—————————
ARTEMISMethod category=Traditional Model-Based Methods, Venue=ICLR'222026.0350.2927.1652.421.7843.6429.254.8326.05—————————
MLLM-I2WBackbone=CLIP-L, # Triplets=3M, Zero-shot=true2025.1150.1——————30.3—————————
CIReVL†Arch=ViT-G/14, Train=false, Pre-trained Weights Source=OpenCLIP (rerun)2026.0550.125.0945.1232.6350.0534.4255.1230.71—————————
SSNNoise=20%2026.0449.8122.6145.5627.8748.5831.8255.2827.43——38.62——————
Context-I2W + RTDBackbone=CLIP ViT-L/14, Integration with RTD=true2024.0649.5——————28.1—————————
LinCIRBackbone=FSC-CLIP ViT-L/14, Integration with RTD=false2024.0649.5——————28.3—————————
i-SEARLEBackbone=CLIP-L, # Params=442M, # Triplets=205K, Zero-shot=true2025.1149.5——————29.2—————————
CIReVLArch=ViT-B/32, Train=false, Pre-trained Weights Source=OpenAI2026.0549.3525.2946.3628.3647.8431.2153.8528.28—————————
SEIZEArch=ViT-B/32, Train=false, Pre-trained Weights Source=OpenAI2026.0549.3525.3746.8429.3847.9732.0754.7828.94—————————
Context-I2WBackbone=CLIP ViT-L/14, Integration with RTD=false2024.0649.1——————27.9—————————
Pic2Word + RTDBackbone=CLIP ViT-L/14, Integration with RTD=true2024.0648.9——————27.6—————————
VaGFeM_Triplet#Triplets=10K, Zero-shot=true, Backbone=BLIP-base2025.1148.8——————28.7—————————