disef
Pytorch implementation of "Diversified in-domain synthesis with efficient fine-tuning for few-shot classification"
MIT License
Overview
This is the official repository for the paper:
Diversified in-domain synthesis with efficient fine-tuning for few-shot classification Victor G. Turrisi da Costa*, Nicola Dall'Asen*, Yiming Wang, Nicu Sebe and Elisa Ricci.
The code is divided into two main parts, one for fine-tuning a pre-trained model
in the few-shot scenario (fine-tune) and the other for generating synthetic data
to enhance fine-tuning (generation).
Each part contains its respective README files available in fine-tune/README.md and generation/README.md with additional details about installation, code organization and execution.
Citation
@misc{dacosta2023diversified,
title={Diversified in-domain synthesis with efficient fine-tuning for few-shot classification},
author={Victor G. Turrisi da Costa and Nicola Dall'Asen and Yiming Wang and Nicu Sebe and Elisa Ricci},
year={2023},
eprint={2312.03046},
archivePrefix={arXiv},
primaryClass={cs.CV}
}