-
We are the first to identify neural vocoders as a source of features to expose synthetic human voices. Here are the differences shown by the six vocoders compared to the original audio:
-
We provide LibriSeVoC as a dataset of self-vocoding samples created with six state-of-the-art vocoders to highlight and exploit the vocoder artifacts. The composition of the data set is shown in the following table:
The source of our dataset ground truth comes from LibriTTS. Therefore, we follow the naming logic of LibriTTS. For example, 27_123349_000006_000000.wav, 27 is the reader's ID, and 123349 is the ID of the chapter.
We propose a new approach to detecting synthetic human voices by exposing signal artifacts left by neural vocoders and modifying and improving the RawNet2 baseline by adding multi-loss, lowering the error rate from 6.10% to 4.54% on the ASVspoof Dataset.
This is the framework of the proposed synthesized voice detection method:
For more details, please read our paper: https://openaccess.thecvf.com/content/CVPR2023W/WMF/html/Sun_AI-Synthesized_Voice_Detection_Using_Neural_Vocoder_Artifacts_CVPRW_2023_paper.html
For more details, please download our dataset: https://drive.google.com/file/d/1NXF9w0YxzVjIAwGm_9Ku7wfLHVbsT7aG/view
python main.py --data_path /your/path/to/LibriSeVoc/ --model_save_path /your/path/to/models/
Test on our Lab's Deepfake O Meter: https://zinc.cse.buffalo.edu/ubmdfl/deep-o-meter/landing_page