# Unveiling Audio Deepfake Origins: A Deep Metric learning And Conformer Network Approach With Ensemble Fusion

Source: [Unveiling Audio Deepfake Origins: A Deep Metric learning And Conformer Network Approach With Ensemble Fusion](<https://www.sota2.com/research/paper/68fa79181e209131cea7c1a7>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T04:52:59.877Z

Authors: Ajinkya Kulkarni, Sandipana Dowerah, Tanel Alumae, Mathew Magimai.-Doss

Published: 2025-06-02T12:42:09.000Z

## Abstract

Audio deepfakes are acquiring an unprecedented level of realism with advanced AI. While current research focuses on discerning real speech from spoofed speech, tracing the source system is equally crucial. This work proposes a novel audio source tracing system combining deep metric multi-class N-pair loss with Real Emphasis and Fake Dispersion framework, a Conformer classification network, and ensemble score-embedding fusion. The N-pair loss improves discriminative ability, while Real Emphasis and Fake Dispersion enhance robustness by focusing on differentiating real and fake speech patterns. The Conformer network captures both global and local dependencies in the audio signal, crucial for source tracing. The proposed ensemble score-embedding fusion shows an optimal trade-off between in-domain and out-of-domain source tracing scenarios. We evaluate our method using Frechet Distance and standard metrics, demonstrating superior performance in source tracing over the baseline system.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Audio Source Attribution on MLAAD In-Domain (Eval)](<https://www.sota2.com/research/sota/audio-source-attribution-on-mlaad-in-domain-eval.md>): Audio Source Attribution; MLAAD In-Domain (Eval); Accuracy: 95.6.

- [Audio Source Attribution on MLAAD (OOD Eval)](<https://www.sota2.com/research/sota/audio-source-attribution-on-mlaad-ood-eval.md>): Audio Source Attribution; MLAAD (OOD Eval); Accuracy: 44.8.

- [OOD Detection on MLAAD v5](<https://www.sota2.com/research/sota/ood-detection-on-mlaad-v5.md>): OOD Detection; MLAAD v5; Accuracy: 44.82.

- [Closed-set Attribution on MLAAD v5](<https://www.sota2.com/research/sota/closed-set-attribution-on-mlaad-v5.md>): Closed-set Attribution; MLAAD v5; Accuracy: 95.61.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2506.02085>)

[PDF](<https://arxiv.org/pdf/2506.02085.pdf>)