Generating Synthetic Datasets by Interpolating along Generalized Geodesics

This is the official Python implementation of the UAI 2023 paper Generating Synthetic Datasets by Interpolating along Generalized Geodesics by Jiaojiao Fan and David Alvarez-Melis.

The repository contains reproducible PyTorch source code for few-shot transfer learning on *NIST and VTAB datasets.

To reproduce results on dataset *NIST

Please go to the folder nist, and refer to the README file there.

cd nist

To reproduce results on dataset VTAB

Please go to the folder vtab, and refer to the README file there.

cd vtab

Citation

If you find this repo useful for your research, please consider citing the paper

@article{fan2023generating,
  title={Generating synthetic datasets by interpolating along generalized geodesics},
  author={Fan, Jiaojiao and Alvarez-Melis, David},
  journal={arXiv preprint arXiv:2306.06866},
  year={2023}
}

For any questions, please contact Jiaojiao Fan (jiaojiaofan@gatech.edu).