Deepfake_Detection

Feature extractor for Pengi(base), CLAP(2023), wav2vec, wavlm

Create Environment

conda create -n "feat_extract" python=3.8
cd models/mspengi
pip install -r requirements.txt

Download Pengi checkpoint

• Download Pengi checkpoint from here
• Move the downloaded checkpoint to AudioEntailment/mspengi/configs/