mlsanityData to 14 Oct 2026

Improving Out-of-Domain Audio Deepfake Detection via Layer Selection and Fusion of SSL-Based Countermeasures

arXiv 2509.12003 ↗First posted 15 Sep 2025cs.SD cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Audio deepfake detection systems based on frozen pre-trained self-supervised learning (SSL) encoders show a high level of performance when combined with layer-weighted pooling methods, such as multi-head factorized attentive pooling (MHFA). However, they still struggle to generalize to out-of-domain (OOD) conditions. We tackle this problem by studying the behavior of six different pre-trained SSLs, on four different test corpora. We perform a layer-by-layer analysis to determine which layers contribute most. Next, we study the pooling head, comparing a strategy based on a single layer with automatic selection via MHFA. We observed that selecting the best layer gave very good results, while reducing system parameters by up to 80%. A wide variation in performance as a function of test corpus and SSL model is also observed, showing that the pre-training strategy of the encoder plays a role. Finally, score-level fusion of several encoders improved generalization to OOD attacks.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 16 of its 23 references that are in the map. First posted on arXiv 15 Sep 2025.

Builds on

16 of 23 references.

References · 16

  1. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
  2. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,657
  3. 2021XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scalecited by 601
  4. 2023Scaling Speech Technology to 1,000+ Languagescited by 411
  5. 2022BEATs: Audio Pre-Training with Acoustic Tokenizerscited by 388
  6. 2021Layer-wise Analysis of a Self-supervised Speech Representation Modelcited by 298
  7. 2022Does Audio Deepfake Detection Generalize?cited by 135
  8. 2022Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentationcited by 117
  9. 2024ASVspoof 5: Crowdsourced Speech Data, Deepfakes, and Adversarial Attacks at Scalecited by 91
  10. 2024MLAAD: The Multi-Language Audio Anti-Spoofing Datasetcited by 56
  11. 2023Towards generalisable and calibrated synthetic speech detection with self-supervised representationscited by 23
  12. 2023Speech Self-Supervised Representations Benchmarking: a Case for Larger Probing Headscited by 17
4 more
  1. 2022An attention-based backend allowing efficient fine-tuning of transformer models for speaker verificationcited by 16
  2. 2024LlamaPartialSpoof: An LLM-Driven Fake Speech Dataset Simulating Disinformation Generationcited by 14
  3. 2025Comprehensive Layer-wise Analysis of SSL Models for Audio Deepfake Detectioncited by 13
  4. 2024Training-Free Deepfake Voice Recognition by Leveraging Large-Scale Pre-Trained Modelscited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Trajectory Dynamics in Self-Supervised Learning Latent Space for Audio Deepfake Detectioncited by 0
  2. 2026Spooftral: Can Voxtral Audio-Language Model Detect Speech Spoofing?cited by 0
  3. 2026GLAD: Global-Local Adaptive Detector for Robust Speech Deepfake Detectioncited by 0