Noisy-Label paper implementations

# noise-validation Some **Noisy-label**, **Robust learning** and **Training tricks** implementation. Code:[https://github.com/ivanwhaf/noise-validation](https://github.com/ivanwhaf/noise-validation) ## Algorithms ### 1.Baseline *MNIST* and *CIFAR10* dataset baseline without label noise. ### 2.QBC-Loss Referring to Active Learning. Using several models to inference, calculating each sample's weighted average loss ### 3.O2U-Net Title: *O2U-Net: A Simple Noisy Label Detection Approach for Deep Neural Networks* Paper: [https://openaccess.thecvf.com/content_ICCV_2019/papers/Huang_O2U-Net_A_Simple_Noisy_Label_Detection_Approach_for_Deep_Neural_ICCV_2019_paper.pdf](https://openaccess.thecvf.com/content_ICCV_2019/papers/Huang_O2U-Net_A_Simple_Noisy_Label_Detection_Approach_for_Deep_Neural_ICCV_2019_paper.pdf) ### 4.Knowledge distilling Title: *Distilling the Knowledge in a Neural Network* Paper: [https://arxiv.org/pdf/1503.02531.pdf](https://arxiv.org/pdf/1503.02531v1.pdf) ### 5.Mean-Teacher Title: *Mean teachers are better role models:Weight-averaged consistency targets improve semi-supervised deep learning results* Paper: [https://arxiv.org/pdf/1703.01780.pdf](https://arxiv.org/pdf/1703.01780v6.pdf) ### 6.Variance Title: *Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samples* Paper: [https://arxiv.org/pdf/1704.07433.pdf](https://arxiv.org/pdf/1704.07433v4.pdf) ### 7.Decoupling Title: *Decoupling “when to update” from “how to update”* Paper: [https://arxiv.org/pdf/1706.02613.pdf](https://arxiv.org/pdf/1706.02613v2.pdf) ### 8.MixUp Title: *mixup: BEYOND EMPIRICAL RISK MINIMIZATION* Paper: [https://arxiv.org/pdf/1710.09412.pdf](https://arxiv.org/pdf/1710.09412v2.pdf) ### 9.MentorNet (Haven't done) Title: *MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels* Paper: [https://arxiv.org/pdf/1712.05055.pdf](https://arxiv.org/pdf/1712.05055v2.pdf) ### 10.Co-teaching Title: *Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels* Paper: [https://arxiv.org/pdf/1804.06872.pdf](https://arxiv.org/pdf/1804.06872v3.pdf) ### 11.Truncated Loss(L_q Loss, GCE) Title: *Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels* Paper: [https://arxiv.org/pdf/1805.07836v4.pdf](https://arxiv.org/pdf/1805.07836v4.pdf) ### 12.Forgetting Title: *AN EMPIRICAL STUDY OF EXAMPLE FORGETTING DURING DEEP NEURAL NETWORK LEARNING* Paper: [https://arxiv.org/pdf/1812.05159.pdf](https://arxiv.org/pdf/1812.05159v3.pdf) ### 13.Meta-Weight-Net (Haven't done) Title: *Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting* Paper: [https://arxiv.org/pdf/1902.07379.pdf](https://arxiv.org/pdf/1902.07379v6.pdf) ### 14.MixMatch Title: *MixMatch: A Holistic Approach to Semi-Supervised Learning* Paper: [https://arxiv.org/pdf/1905.02249v1.pdf](https://arxiv.org/pdf/1905.02249v1.pdf) ### 15.SCE Title: *Symmetric Cross Entropy for Robust Learning with Noisy Labels* Paper: [https://arxiv.org/pdf/1908.06112.pdf](https://arxiv.org/pdf/1908.06112v1.pdf) ### 16.NLNL Title: *NLNL: Negative Learning for Noisy Labels* Paper: [https://arxiv.org/pdf/1908.07387.pdf](https://arxiv.org/pdf/1908.07387v1.pdf) ### 17.SELF Title: *SELF: LEARNING TO FILTER NOISY LABELS WITH SELF-ENSEMBLING* Paper: [https://arxiv.org/pdf/1910.01842.pdf](https://arxiv.org/pdf/1910.01842v1.pdf) ### 18.DivideMix (Haven't done) Title: *DIVIDEMIX: LEARNING WITH NOISY LABELS AS SEMI-SUPERVISED LEARNING* Paper: [https://arxiv.org/pdf/2002.07494v1.pdf](https://arxiv.org/pdf/2002.07494v1.pdf) ### 19.Flooding Title: *Do We Need Zero Training Loss After Achieving Zero Training Error?* Paper: [https://arxiv.org/pdf/2002.08709.pdf](https://arxiv.org/pdf/2002.08709v1.pdf) ### 20.APL Title *Normalized Loss Functions for Deep Learning with Noisy Labels* Paper: [https://arxiv.org/pdf/2006.13554v1.pdf](https://arxiv.org/pdf/2006.13554v1.pdf) ### Label Smoothing ### CE-MAE