A Machine Learning project aimed for a countermeasures of speech spoofing 🤖.
A basic lecture video for self supervised learning.
Download ASVspoof 2019 dataset from here. Run ⚡️
pip install -r requirements.txt
in your terminal to install all required packages.
wave2vec2_representation.py is to get feature embedding by wave2vec 2.0 pretrained model. Adjust the file/directory path to read from train/validation/test dataset and write to designated csv file. Run this script after navigated to the directory of audio files (possible three directories for train/dev/eval).
ml_models.ipynb contains the implementation of all the ML models on the embedded reprsentation of ASVspoof 2019 dataset.
flop_count_2.py computes GMAC.
If you find our work helpful in your research please cite our paper by 🔗
@misc{saha2024exploringgreenaiaudio,
title={Exploring Green AI for Audio Deepfake Detection},
author={Subhajit Saha and Md Sahidullah and Swagatam Das},
year={2024},
eprint={2403.14290},
archivePrefix={arXiv},
primaryClass={cs.SD},
url={https://arxiv.org/abs/2403.14290},
}