CodecFake: Enhancing Anti-Spoofing Models Against Deepfake Audios from Codec-Based Speech Synthesis Systems

Paper, Dataset, Project Page

Interspeech 2024

TL;DR: We show that better detection of deepfake speech from codec-based TTS systems can be achieved by training models on speech re-synthesized with neural audio codecs. We also release the CodecFake dataset for this purpose.

Dataset Download

We provide the CodecFake dataset in two forms:

1. Huggingface Datasets

   from datasets import load_dataset
   a = load_dataset("rogertseng/CodecFake")

2. ZIP files

Train Fake Speech Detectors on CodecFake

See instructions under detection for more.

Dataset Creation Pipeline

TBA, see dataset_creation

Acknowledgement

CodecFake is created based on the VCTK dataset, licensed under CC-BY-4.0.