Detail výsledku
Codec-Based Deepfake Source Tracing via Neural Audio Codec Taxonomy
Lin I. Ming
Zhang Lin, Ph.D.
Du Jiawei
Wu Haibin
Lee Hung Yi
Jang Jyh Shing Roger
Recent advances in neural audio codec-based speech generation (CoSG) models have produced remarkably realistic audio deepfakes. We refer to deepfake speech generated by CoSG systems as codec-based deepfake, or CodecFake. Although existing anti-spoofing research on CodecFake predominantly focuses on verifying the authenticity of audio samples, almost no attention was given to tracing the CoSG used in generating these deepfakes. In CodecFake generation, processes such as speech-to-unit encoding, discrete unit modeling, and unit-to-speech decoding are fundamentally based on neural audio codecs. Motivated by this, we introduce source tracing for CodecFake via neural audio codec taxonomy, which dissects neural audio codecs to trace CoSG. Our experimental results on the CodecFake+ dataset provide promising initial evidence for the feasibility of CodecFake source tracing while also highlighting several challenges that warrant further investigation.
Anti-spoofing | audio deepfake detection | explainability | neural audio codec | source tracing
@inproceedings{BUT199995,
author="{} and {} and Lin {Zhang} and {} and {} and {} and {}",
title="Codec-Based Deepfake Source Tracing via Neural Audio Codec Taxonomy",
booktitle="Proceedings of the Annual Conference of the International Speech Communication Association Interspeech",
year="2025",
journal="Interspeech",
pages="1538--1542",
publisher="International Speech Communication Association",
address="Rotterdam, Nizozemí",
doi="10.21437/Interspeech.2025-1297",
url="https://www.isca-archive.org/interspeech_2025/chen25j_interspeech.pdf"
}