REVIEW 3 cited by
Investigating Prosodic Signatures via Speech Pre-Trained Models for Audio Deepfake Source Attribution
Not yet reviewed by Pith; the record is open.
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
read the original abstract
In this work, we investigate various state-of-the-art (SOTA) speech pre-trained models (PTMs) for their capability to capture prosodic signatures of the generative sources for audio deepfake source attribution (ADSD). These prosodic characteristics can be considered one of major signatures for ADSD, which is unique to each source. So better is the PTM at capturing prosodic signs better the ADSD performance. We consider various SOTA PTMs that have shown top performance in different prosodic tasks for our experiments on benchmark datasets, ASVSpoof 2019 and CFAD. x-vector (speaker recognition PTM) attains the highest performance in comparison to all the PTMs considered despite consisting lowest model parameters. This higher performance can be due to its speaker recognition pre-training that enables it for capturing unique prosodic characteristics of the sources in a better way. Further, motivated from tasks such as audio deepfake detection and speech recognition, where fusion of PTMs representations lead to improved performance, we explore the same and propose FINDER for effective fusion of such representations. With fusion of Whisper and x-vector representations through FINDER, we achieved the topmost performance in comparison to all the individual PTMs as well as baseline fusion techniques and attaining SOTA performance.
Forward citations
Cited by 3 Pith papers
-
Open-Set Source Tracing of Audio Deepfake Systems
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.
-
Attention Is Not Always the Answer: Optimizing Voice Activity Detection with Simple Feature Fusion
FusionVAD shows that simple addition or concatenation of MFCC and pre-trained model features outperforms cross-attention fusion for voice activity detection, with the best model beating Pyannote by 2.04 average DER.
-
Source Tracing of Synthetic Speech Systems Through Paralinguistic Pre-Trained Representations
Paralinguistic speech representations, especially TRILLsson, are the most effective single features for tracing synthetic speech to its source generator, and the TRIO fusion with x-vector reports new accuracy highs.
Discussion (0). Sign in to comment.