Co-teaching (Inception-ResNet-v2)

Reported on 4 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision4 results

Image Classificationonmini WebVision 1.0
Top-1 Accuracy· 2018-04-18
63.58
best: 84.16 (LRA-diffusion (CLIP ViT))
SOTA
Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels arXiv:1804.06872
Image Classificationonmini WebVision 1.0
Top-5 Accuracy· 2018-04-18
85.2
best: 94.84 (PSSCL (130 epochs))
SOTA
Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels arXiv:1804.06872
Image Classificationonmini WebVision 1.0
ImageNet Top-1 Accuracy· 2018-04-18
61.48
best: 82.56 (LRA-diffusion (CLIP ViT))
Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels arXiv:1804.06872
Image Classificationonmini WebVision 1.0
ImageNet Top-5 Accuracy· 2018-04-18
84.7
best: 97.24 (RTE (Inception-ResNet-v2))
Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels arXiv:1804.06872