This repository contains our implementation of the paper "Prompt_Tuning_for_Audio_Deepfake_Detection". This work proposed an ADD method using prompt tuning in a plug-in style.It bridges domain gaps by integrating it seamlessly with state-of-the-art transformer models and other fine-tuning methods, boosting their performance on target data. This study accepted at INTERSPEECH 2024
This repository consists of W2V_prompt_tuning based on the wav2vec-based audio deepfake detection method by Tak Hemlata et al. and WSP_prompt_tuning based on the whisper-based audio deepfake detection method by Piotr Kawa et al.
Note that the environments of these two repositories are incompatible.