List your company

Image Classification on ImageNet to 10 Target Datasets (test)

72.33ImageNet AccuracyDeKgTCP
66.495668.010369.52571.0397Oct 6, 2022May 12, 2023Dec 16, 2023Jul 21, 2024Feb 24, 2025Sep 30, 2025May 6, 2026

Evaluation results

MethodLinks
DeKgTCPSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1272.3394.7390.0265.4972.3986.5925.0567.1944.4751.3768.7866.61
TAPSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1272.394.390.765.670.9386.124.5768.350.24668.966.56
IPOShots=16-shot, Backbone=ViT-B/162024.1072.1594.3490.9666.172.7586.7525.1467.9747.0148.5669.2367.36
CoOp + SpecPLSource Dataset=ImageNet2026.0572.1294.1290.6263.6671.1386.0923.8565.7145.0453.5467.3866.11
MMRL + SpecPLSource Dataset=ImageNet2026.0572.0794.8591.9966.4872.9686.8626.5867.9446.4352.8869.167.61
MMRL2025.0372.0394.6791.4366.172.7786.426.367.5745.953.168.2767.25
MMRLSource Dataset=ImageNet2026.0572.0394.6791.4366.172.7786.426.367.5745.953.168.2767.25
TransAgentmode=few-shot prompt learning2024.107294.3790.3365.4371.486.4723.266.245.352.1369.9366.48
DroPLeSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1271.9494.7791.0666.1872.3786.1726.3667.3347.4552.6368.4667.28
HPT++2024.0871.8194.0292.1665.5572.4386.3428.668.7851.0250.7670.5368.02
CDCTransfer Source=ImageNet2024.1071.7694.4790.7766.2772.6786.2724.568.0746.649.1368.666.73
HPTtraining_shots=16, protocol=cross-dataset evaluation2023.1271.7294.292.6366.3374.8486.2125.6868.7550.8747.3670.567.74
HPT2024.0871.7294.292.6366.3374.8486.2125.6868.7550.8747.3670.567.74
CoOpevaluation_mode=cross-dataset evaluation, source_dataset=ImageNet2022.1071.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOptraining_shots=16, protocol=cross-dataset evaluation2023.1271.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOp2023.1171.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOp2024.0871.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOpShots=16-shot, Backbone=ViT-B/162024.1071.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOpTransfer Source=ImageNet2024.1071.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOpmode=few-shot prompt learning2024.1071.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOp2025.0371.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOpSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1271.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
CoOpSource Dataset=ImageNet2026.0571.5193.789.1464.5168.7185.318.4764.1541.9246.3966.5563.88
TCP2025.0371.493.9791.2564.6971.2186.6923.4567.1544.3551.4568.7366.29
ProMetaRSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1271.2993.7490.5965.8371.1386.3924.7867.4147.0845.0269.566.15
PromptSRC2023.1171.2793.690.2565.770.2586.1523.967.146.8745.568.7565.81
PromptSRC2024.0871.2793.690.2565.770.2586.1523.967.146.8745.568.7565.81
PromptSRCmode=few-shot prompt learning2024.1071.2793.690.2565.770.2586.1523.967.146.8745.568.7565.81
PromptSRC2025.0371.2793.690.2565.770.2586.1523.967.146.8745.568.7565.81
PSRCSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1271.2793.690.2565.770.2586.1523.967.146.8745.568.7565.81
Co-CoOpevaluation_mode=cross-dataset evaluation, source_dataset=ImageNet2022.1071.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
CoCoOptraining_shots=16, protocol=cross-dataset evaluation2023.1271.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
CoCoOp2023.1171.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
CoCoOp2024.0871.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
CoCoOpShots=16-shot, Backbone=ViT-B/162024.1071.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
Co-CoOpTransfer Source=ImageNet2024.1071.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
CoCoOpmode=few-shot prompt learning2024.1071.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
Co-CoOp2025.0371.0294.4390.1465.3271.8886.0622.9467.3645.7345.3768.2165.74
MMA2025.037193.890.366.1372.0786.1225.3368.1746.5749.2468.3266.61
TEntype=Tuning ensemble2023.1170.8893.9190.7271.9472.5986.6826.0366.0749.3148.1869.1467.46
CoPrompt2024.0870.894.590.7365.6772.386.432467.5747.0751.969.7367
CoPromptSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1270.894.590.7365.6772.386.432467.5747.0751.969.7367
MaPLeevaluation_mode=cross-dataset evaluation, source_dataset=ImageNet2022.1070.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
MaPLetraining_shots=16, protocol=cross-dataset evaluation2023.1270.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
MaPLe2023.1170.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
MaPLe2024.0870.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
MaPLeTransfer Source=ImageNet2024.1070.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
MaPLemode=few-shot prompt learning2024.1070.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
MaPLe2025.0370.7293.5390.4965.5772.2386.224.7467.0146.4948.0668.6966.3
KgCoOpSource Dataset=ImageNet, Few-shot Setting (Shots)=16 shots2025.1270.6693.9289.8365.4170.0186.3622.5166.1646.3546.0468.565.51
CLIP2023.1166.7292.9489.0765.2971.386.1124.8762.6244.5647.6966.7765.12
CLIPmode=zero-shot2024.1066.7292.9489.0765.2971.386.1124.8762.6244.5647.6966.7765.12