Label Noise

• large negative loss rejection
  • [2206.03740] Large Loss Matters in Weakly Supervised Multi-Label Classification

Robust Losses

• DiscrimLoss: A Universal Loss for Hard Samples and Incorrect Samples Discrimination (2022-08-22)

Curriculum Learning

Noise Correction

• GitHub - MrChenFeng/SSR_BMVC2022: SSR: An Efficient and Robust Framework for Learning with Unknown Label Noise
• GitHub - Justherozen/ProMix: PyTorch Code for ProMix: Combating Label Noise via Maximizing Clean Sample Utility

Early Stopping

GitHub - mahf93/tracking-memorization: Code for “Leveraging Unlabeled Data to Track Memorization”

Datasets

• GitHub - allegro/AlleNoise