List your company
ResearchFormerly WizWand

Image Classification on MNIST Label Noise 0.2 (test)

94.6Mean AccuracyGIW
78.313682.541886.7790.9982May 24, 2023
Updated 8mo ago

Evaluation results

MethodLinks
GIWShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0594.6
GIWShift Scenario=Case (iv)2023.0591.44
ReweightShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0589.28
CCSAShift Scenario=Case (iv)2023.0588.06
Pretrain-valShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0586.34
DANNShift Scenario=Case (iv)2023.0585.32
Pretrain-valShift Scenario=Case (iv)2023.0585.08
R-DIWShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0582.7
MW-NetShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0582.29
DIWShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0580.12
Val-onlyShift Scenario=Case (iv)2023.0579.86
Val-onlyShift type=Label Noise, Shift intensity=0.2, Case=iii, Trials=52023.0578.94