Home

Awesome

Contrastive-Finetuning

This repo is the official implementation of the following paper: "On the Importance of Distractors for Few-Shot Classification" Paper

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

@misc{das2021importance,
    title={On the Importance of Distractors for Few-Shot Classification},
    author={Rajshekhar Das and Yu-Xiong Wang and JoséM. F. Moura},
    year={2021},
    eprint={2109.09883},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

Dataset Download

To set up the dataset, follow the exact steps outlined in here.

Pretrained Model

To download the pretrained backbone model, follow the exact steps outlined in here

Running

Acknowlegements

Part of the codebase, namely, the dataloaders have been adapted from Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation.