First, clone the repository locally, create and activate a conda environment, and install the requirements :
$ git clone https://github.com/TakHemlata/SSL_Anti-spoofing.git
$ conda create -n SSL_Spoofing python=3.7
$ conda activate SSL_Spoofing
$ pip install torch==1.8.1+cu111 torchvision==0.9.1+cu111 torchaudio==0.8.1 -f https://download.pytorch.org/whl/torch_stable.html
$ cd fairseq-a54021305d6b3c4c5959ac9395135f63202db8f1
(This fairseq folder can also be downloaded from https://github.com/pytorch/fairseq/tree/a54021305d6b3c4c5959ac9395135f63202db8f1)
$ pip install --editable ./
$ pip install -r requirements.txt
Our experiments are trained on 2019 LA training and evaluated on 2015 evaluation,2019 LA evaluation,2021 LA and DF evaluation and in the wild evaluation database.
Download the XLSR models from here
Please change the address in Need to modify the path.txt first.
cd myresult
python test_dual.py
The complete code of the subsequent training will be added after the successful publication of the paper. Thank you for your attention.