Interleaving clean and noisy training epochs improves clean, corrupted, and out-of-distribution accuracy on CIFAR-100 and ImageNet for CNNs and ViTs, with impulse noise best for ResNets and Gaussian noise best for ViTs.
Enzo Marinari and Giorgio Parisi
1 Pith paper cite this work, alongside 16 external citations. Polarity classification is still indexing.
1
Pith paper citing it
16
external citations · OpenAlex
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Interleaved Noise Injection Improves Clean, Corrupted, and OOD Performance
Interleaving clean and noisy training epochs improves clean, corrupted, and out-of-distribution accuracy on CIFAR-100 and ImageNet for CNNs and ViTs, with impulse noise best for ResNets and Gaussian noise best for ViTs.