/symmetric_cross_entropy_for_noisy_labels

Code for ICCV2019 "Symmetric Cross Entropy for Robust Learning with Noisy Labels"

Primary LanguagePython

Symmetric Learning (SL) via Symmetric Cross Entropy (SCE) loss

Code for ICCV2019 "Symmetric Cross Entropy for Robust Learning with Noisy Labels" https://arxiv.org/abs/1908.06112

Requirements

  • Python 3.5.2
  • Tensorflow 1.10.1
  • Keras 2.2.2

Usage

Simply run the code by python3 train_models.py

It can config with dataset, model, epoch, batchsize, noise_rate, symmetric or asymmetric type noise

The Pytorch reimplementation

The Pytorch version is implemented by Hanxun Huang. The code can be found here: https://github.com/HanxunHuangLemonBear/SCELoss-Reproduce

Citing this work

If you use this code in your work, please cite the accompanying paper:

@inproceedings{wang2019symmetric,
  title={Symmetric cross entropy for robust learning with noisy labels},
  author={Wang, Yisen and Ma, Xingjun and Chen, Zaiyi and Luo, Yuan and Yi, Jinfeng and Bailey, James},
  booktitle={IEEE International Conference on Computer Vision},
  year={2019}
}