Pith. sign in

REVIEW 4 cited by

Neural Codec Source Tracing: Toward Comprehensive Attribution in Open-Set Condition

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2501.06514 v1 pith:XXCUPF6P submitted 2025-01-11 cs.SD cs.AIeess.AS

Yuankun Xie , Xiaopeng Wang , Zhiyong Wang , Ruibo Fu , Zhengqi Wen , Songjun Cao , Long Ma , Chenxing Li
show 2 more authors
Haonnan Cheng Long Ye
This is my paper
classification cs.SDcs.AIeess.AS
keywords sourcetracingaudiocodecncstneuralopen-setclassification
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Current research in audio deepfake detection is gradually transitioning from binary classification to multi-class tasks, referred as audio deepfake source tracing task. However, existing studies on source tracing consider only closed-set scenarios and have not considered the challenges posed by open-set conditions. In this paper, we define the Neural Codec Source Tracing (NCST) task, which is capable of performing open-set neural codec classification and interpretable ALM detection. Specifically, we constructed the ST-Codecfake dataset for the NCST task, which includes bilingual audio samples generated by 11 state-of-the-art neural codec methods and ALM-based out-ofdistribution (OOD) test samples. Furthermore, we establish a comprehensive source tracing benchmark to assess NCST models in open-set conditions. The experimental results reveal that although the NCST models perform well in in-distribution (ID) classification and OOD detection, they lack robustness in classifying unseen real audio. The ST-codecfake dataset and code are available.

Discussion (0). Sign in to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Towards Neural Audio Codec Source Parsing

    eess.AS 2025-06 conditional novelty 7.0 of 10

    NACSP predicts codec generation parameters from deepfake audio, and the proposed HYDRA hyperbolic model beats Euclidean baselines on most benchmark tasks.

  2. Multilingual Source Tracing of Speech Deepfakes: A First Benchmark

    eess.AS 2025-08 conditional novelty 6.0 of 10

    The first multilingual source-tracing benchmark for speech deepfakes, showing LFCC-ECAPA-TDNN generalizes best across languages.

  3. Open-Set Source Tracing of Audio Deepfake Systems

    eess.AS 2025-07 conditional novelty 5.0 of 10

    Softmax energy, a modified out-of-distribution score, improves open-set source tracing of audio deepfake systems, achieving a 31% relative FPR95 reduction and best FPR95 of 8.3% with augmentation.

  4. Towards Generalized Source Tracing for Codec-Based Deepfake Speech

    cs.SD 2025-06 conditional novelty 5.0 of 10

    SASTNet, which fuses Whisper semantic features with Wav2Vec2 and AudioMAE acoustic features, improves source tracing for codec-based deepfake speech on CodecFake+, while exposing that prior models overfit to silence.

Pith tools