List your company
ResearchFormerly WizWand

Image Classification on CIFAR-100-LTN Asymmetric Noise IF=10, NR=40% (test)

70.5AccuracyCLIP+CARE
26.8238.1649.560.84May 22, 2026
Updated 5mo ago

Evaluation results

MethodLinks
CLIP+CAREBackbone=CLIP, Noise type=Asymmetric, Imbalance Factor (IF)=10, Noise Rate (NR)=40%2026.0570.5
CLIP+LA w. CAREFoundation Model=CLIP2026.0570.5
CLIP+IBCBackbone=CLIP, Noise type=Asymmetric, Imbalance Factor (IF)=10, Noise Rate (NR)=40%2026.0570.2
CLIP+RLA*Foundation Model=CLIP2026.0569.3
CLIP+LAFoundation Model=CLIP2026.0568.2
CLIP+CE w. CAREFoundation Model=CLIP2026.0567.5
CLIPProtocol=zero-shot2026.0564.4
CLIP+CEFoundation Model=CLIP2026.0564.4
ECBS2026.0552.1
TABASCO2026.0550.5
UNICON2026.0544.7
ULC2026.0543.2
DivideMix2026.0542
RoLT2026.0539.3
IB2026.0535.3
LDAM2026.0533.3
HAR2026.0533.2
CE2026.0532.1
MW-Net2026.0530.4
LA2026.0528.5