Skip to content

Code for Data-Efficient Augmentation for Training Neural Networks (NeurIPS 2022)

License

Notifications You must be signed in to change notification settings

tianyu139/data-efficient-augmentation

Repository files navigation

Data-Efficient Augmentation for Training Neural Networks (NeurIPS 2022)

DEA

Official code repository for Data-Efficient Augmentation for Training Neural Networks.

Authors: Tian Yu Liu and Baharan Mirzasoleiman

Datasets

Create a data/ folder in the main directory and place datasets there (e.g. Caltech256, TinyImageNet, ImageNet)

Example Usage

Training on and augmenting 30% coresets on Caltech256

bash train.sh

Training on random and augmenting 50% coresets on ImageNet

bash train_imagenet.sh 

Acknowledgements

If you find this useful for your work, please consider citing

@article{liu2022data,
  title={Data-Efficient Augmentation for Training Neural Networks},
  author={Liu, Tian Yu and Mirzasoleiman, Baharan},
  journal={Advances in Neural Information Processing Systems (NeurIPS)},
  year={2022}
}

About

Code for Data-Efficient Augmentation for Training Neural Networks (NeurIPS 2022)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published