🎧 CodecFake+: A Large-Scale Neural Audio Codec–Based Deepfake Speech Dataset

A benchmark for studying neural codec–based deepfake speech detection.

---

🧩 Overview

CodecFake+ is a large-scale benchmark designed to evaluate the detection and generalization capabilities of models across diverse neural audio codec–generated deepfakes.
It enables controlled, taxonomy-aware analysis for codec-based deepfake research.

---

📰 News

🚧 The CodecFake+ project page is currently under construction.
Thank you for your patience and continued interest.

🔗 Official webpage: responsiblegenai.github.io/CodecFake-Plus-Dataset

Citations

If you find our dataset useful, please consider citing our accompanying papers.

@article{chen2025codecfake+,
  title   = {CodecFake+: A Large-Scale Neural Audio Codec-Based Deepfake Speech Dataset},
  author  = {Chen, Xuanjun and Du, Jiawei and Wu, Haibin and Zhang, Lin and Lin, I and Chiu, I and Ren, Wenze and Tseng, Yuan and Tsao, Yu and Jang, Jyh-Shing Roger and others},
  journal = {arXiv preprint arXiv:2501.08238},
  year    = {2025}
}

@inproceedings{wu24p_interspeech,
  title     = {{CodecFake: Enhancing Anti-Spoofing Models Against Deepfake Audios from Codec-Based Speech Synthesis Systems}},
  author    = {Haibin Wu and Yuan Tseng and Hung-yi Lee},
  year      = {2024},
  booktitle = {{Interspeech 2024}},
  pages     = {1770--1774},
  doi       = {10.21437/Interspeech.2024-2093},
  issn      = {2958-1796},
}