List your company
ResearchFormerly WizWand

Classification on STL10 Noise (test)

98.58Accuracy (Original)EC
64.5273.362582.20591.0475May 16, 2025
Updated 6mo ago

Evaluation results

MethodLinks
ECClassifier Type=End-to-end (EC), Features=DINOv22025.0598.5858.22
MaWaClassifier Type=Nonlinear (NC), Features=DINOv22025.0598.5586.23
MaWaClassifier Type=Linear (LC), Features=DINOv22025.0598.1186.09
WaCoClassifier Type=Nonlinear (NC), Features=DINOv22025.0597.7180.75
WaCoClassifier Type=Linear (LC), Features=DINOv22025.0597.2678.78
HSICClassifier Type=Nonlinear (NC), Features=DINOv22025.0597.0245.78
HSICClassifier Type=Linear (LC), Features=DINOv22025.0595.5545.47
SCLRClassifier Type=Nonlinear (NC), Features=DINOv22025.0595.457.75
SCLRClassifier Type=Linear (LC), Features=DINOv22025.0594.2444.47
ECEvaluation Protocol=EC, Pre-trained Features=SwAV2025.0594.2340.82
MaWaEvaluation Protocol=LC, Pre-trained Features=SwAV2025.0593.9773.18
MaWaEvaluation Protocol=NC, Pre-trained Features=SwAV2025.0593.5972.84
WaCoEvaluation Protocol=LC, Pre-trained Features=SwAV2025.0587.1768.23
WaCoEvaluation Protocol=NC, Pre-trained Features=SwAV2025.0586.667.7
SCLREvaluation Protocol=NC, Pre-trained Features=SwAV2025.0576.3947.06
HSICEvaluation Protocol=NC, Pre-trained Features=SwAV2025.0574.3865.05
SCLREvaluation Protocol=LC, Pre-trained Features=SwAV2025.0573.9446.88
HSICEvaluation Protocol=LC, Pre-trained Features=SwAV2025.0565.8359.97