mlsanityData to 7 Oct 2026

From Self-Supervised Speech Models to Mixture-of-Experts for Robust Anti-Spoofing

arXiv 2606.14639 ↗First posted 12 Jun 2026cs.SD cs.AIcited by 0

On the map: Mixture-of-experts → Multimodal MoE models

Abstract (arXiv)

Recent advances in speech generation have significantly improved the naturalness of synthetic speech, making spoofing detection increasingly challenging. A key limitation of current anti-spoofing systems is their limited robustness to unseen synthesis methods. In this work, we transform a self-supervised speech representation model into a Mixture-of-Experts (MoE) architecture to improve generalization. Feed-forward blocks in selected encoder layers are replaced by multiple expert networks controlled by a layer-wise gating mechanism, allowing experts to capture complementary acoustic patterns while preserving the representations learned during self-supervised pretraining. We further analyze the architectural choices affecting the performance of this MoE conversion and investigate the activation behavior of the experts. The proposed approach is evaluated on 14 spoofing datasets and reduces the macro EER from 5.46% to 4.81%, corresponding to 11.9% relative improvement over the baseline.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Multimodal MoE models, one of 12 directions in Mixture-of-experts. Builds on 35 of its 41 references that are in the map. First posted on arXiv 12 Jun 2026.

Builds on

35 of 41 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Attention-based Mixture of Experts for Robust Speech Deepfake Detection22 Sep
  2. 2025Wav2DF-TSL: Two-stage Learning with Efficient Pre-training and Hierarchical Experts Fusion for Robust Audio Deepfake Detection4 Sep
  3. 2024UME: Upcycling Mixture-of-Experts for Scalable and Efficient Automatic Speech Recognition23 Dec
  4. 2024Leveraging Mixture of Experts for Improved Speech Deepfake Detection24 Sep
  5. 2024Mixture of Experts Fusion for Fake Audio Detection Using Frozen wav2vec 2.018 Sep

Other references · 30

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  2. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,853
  3. 2021Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsitycited by 3,014
  4. 2021HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Unitscited by 2,301
  5. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,609
  6. 2023Neural Codec Language Models are Zero-Shot Text to Speech Synthesizerscited by 737
  7. 2021Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speechcited by 445
  8. 2022Sparse Upcycling: Training Mixture-of-Experts from Dense Checkpointscited by 207
  9. 2021AASIST: Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networkscited by 173
  10. 2020End-to-end anti-spoofing with RawNet2cited by 160
  11. 2021ASVspoof 2021: accelerating progress in spoofed and deepfake speech detectioncited by 157
  12. 2023Matcha-TTS: A fast TTS architecture with conditional flow matchingcited by 142
18 more
  1. 2022Does Audio Deepfake Detection Generalize?cited by 129
  2. 2022ADD 2022: the First Audio Deep Synthesis Detection Challengecited by 94
  3. 2025A Comprehensive Survey of Mixture-of-Experts: Algorithms, Theory, and Applicationscited by 91
  4. 2024ASVspoof 5: Crowdsourced Speech Data, Deepfakes, and Adversarial Attacks at Scalecited by 88
  5. 2021End-to-End Spectro-Temporal Graph Attention Networks for Speaker Verification Anti-Spoofing and Speech Deepfake Detectioncited by 68
  6. 2023ADD 2023: the Second Audio Deepfake Detection Challengecited by 64
  7. 2021ASVspoof 2019: spoofing countermeasures for the detection of synthesized, converted and replayed speechcited by 56
  8. 2024MLAAD: The Multi-Language Audio Anti-Spoofing Datasetcited by 55
  9. 2023Audio Deepfake Detection: A Surveycited by 49
  10. 2023AI-Synthesized Voice Detection Using Neural Vocoder Artifactscited by 41
  11. 2024The Codecfake Dataset and Countermeasures for the Universally Detection of Deepfake Audiocited by 31
  12. 2018Interpretable Convolutional Filters with SincNetcited by 20
  13. 2024DFADD: The Diffusion and Flow-Matching Based Audio Deepfake Datasetcited by 20
  14. 2025Speech DF Arena: A Leaderboard for Speech DeepFake Detection Modelscited by 20
  15. 2022An attention-based backend allowing efficient fine-tuning of transformer models for speaker verificationcited by 14
  16. 2025MoLEx: Mixture of LoRA Experts in Speech Self-Supervised Models for Audio Deepfake Detectioncited by 5
  17. 2025Adaptive Mixture of Low-Rank Experts for Robust Audio Spoofing Detectioncited by 3
  18. 2025Mixture of Low-Rank Adapter Experts in Generalizable Audio Deepfake Detectioncited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026DFKI-Speech System for WildSpoof Challenge: A robust framework for SASV In-the-Wild
  2. 2024BIG-MoE: Bypass Isolated Gating MoE for Generalized Multimodal Face Anti-Spoofing
  3. 2025Exploiting the Experts: Unauthorized Compression in MoE-LLMs
  4. 2025Beyond Hard Sharing: Efficient Multi-Task Speech-to-Text Modeling with Supervised Mixture of Experts
  5. 2025Omni-Router: Sharing Routing Decisions in Sparse Mixture-of-Experts for Speech Recognition
  6. 2025Speech Quality Assessment Model Based on Mixture of Experts: System-Level Performance Enhancement and Utterance-Level Challenge Analysis
  7. 2022Adaptive Mixture of Experts Learning for Generalizable Face Anti-Spoofing
  8. 2024Differentially Private Training of Mixture of Experts Models