List your company
ResearchFormerly WizWand

Image Classification on MNIST Label Noise 0.4 (test)

93.74Mean AccuracyGIW
56.591266.235675.8885.5244May 24, 2023
Updated 8mo ago

Evaluation results

MethodLinks
GIWShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0593.74
GIWShift Scenario=Case (iv)2023.0588.84
Pretrain-valShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0585.43
ReweightShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0584.73
CCSAShift Scenario=Case (iv)2023.0584
Pretrain-valShift Scenario=Case (iv)2023.0583.26
R-DIWShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0581.19
MW-NetShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0580.92
Val-onlyShift Scenario=Case (iv)2023.0579.86
DIWShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0579
Val-onlyShift type=Label Noise, Shift intensity=0.4, Case=iii, Trials=52023.0578.94
DANNShift Scenario=Case (iv)2023.0558.02