mlsanityData to 7 Oct 2026

A Data-Driven Diffusion-based Approach for Audio Deepfake Explanations

arXiv 2506.03425 ↗First posted 3 Jun 2025eess.AS cs.AI cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Evaluating explainability techniques, such as SHAP and LRP, in the context of audio deepfake detection is challenging due to lack of clear ground truth annotations. In the cases when we are able to obtain the ground truth, we find that these methods struggle to provide accurate explanations. In this work, we propose a novel data-driven approach to identify artifact regions in deepfake audio. We consider paired real and vocoded audio, and use the difference in time-frequency representation as the ground-truth explanation. The difference signal then serves as a supervision to train a diffusion model to expose the deepfake artifacts in a given vocoded audio. Experimental results on the VocV4 and LibriSeVoc datasets demonstrate that our method outperforms traditional explainability techniques, both qualitatively and quantitatively.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 29 of its 33 references that are in the map. First posted on arXiv 3 Jun 2025.

Builds on

29 of 33 references.

References · 29

  1. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,690
  2. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  3. 2018ESRGAN: Enhanced Super-Resolution Generative Adversarial Networkscited by 1,250
  4. 2018WaveGlow: A Flow-based Generative Network for Speech Synthesiscited by 308
  5. 2021SegDiff: Image Segmentation with Diffusion Probabilistic Modelscited by 279
  6. 2020A survey of loss functions for semantic segmentationcited by 216
  7. 2022Does Audio Deepfake Detection Generalize?cited by 129
  8. 2024AttnLRP: Attention-Aware Layer-Wise Relevance Propagation for Transformerscited by 116
  9. 2022Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentationcited by 110
  10. 2021WaveFake: A Data Set to Facilitate Audio Deepfake Detectioncited by 78
  11. 2021RawBoost: A Raw Data Boosting and Augmentation Method applied to Automatic Speaker Verification Anti-Spoofingcited by 64
  12. 2024Common Sense Reasoning for Deepfake Detectioncited by 64
17 more
  1. 2024FakeInversion: Learning to Detect Images from Unseen Text-to-Image Models by Inverting Stable Diffusioncited by 61
  2. 2019Neural source-filter waveform models for statistical parametric speech synthesiscited by 46
  3. 2023AI-Synthesized Voice Detection Using Neural Vocoder Artifactscited by 41
  4. 2024DiffusionFake: Enhancing Generalization in Deepfake Detection via Guided Stable Diffusioncited by 29
  5. 2024A Survey on Speech Deepfake Detectioncited by 26
  6. 2022Spoofed training data for speech spoofing countermeasure can be efficiently created using neural vocoderscited by 19
  7. 2023Can large-scale vocoded spoofed data improve speech spoofing countermeasure with a self-supervised front end?cited by 17
  8. 2020An explainability study of the constant Q cepstral coefficient spoofing countermeasure for automatic speaker verificationcited by 13
  9. 2021Explaining deep learning models for spoofing and deepfake detection with SHapley Additive exPlanationscited by 13
  10. 2024Listenable Maps for Audio Classifierscited by 13
  11. 2024How Do Neural Spoofing Countermeasures Detect Partially Spoofed Audio?cited by 10
  12. 2022Explainable deepfake and spoofing detection: an attack analysis using SHapley Additive exPlanationscited by 5
  13. 2023Diffusion-Based Adversarial Purification for Speaker Verificationcited by 4
  14. 2024Interpretable Temporal Class Activation Representation for Audio Spoofing Detectioncited by 4
  15. 2024LMAC-TD: Producing Time Domain Explanations for Audio Classifierscited by 4
  16. 2025What Does an Audio Deepfake Detector Focus on? A Study in the Time Domaincited by 4
  17. 2023Diff-SV: A Unified Hierarchical Framework for Noise-Robust Speaker Verification Using Score-Based Diffusion Probabilistic Modelscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026The Perceived Fragility of Explanations in Audio Models: Manipulation of Attribution with Unchanged Predictionscited by 0
  2. 2026Audio-Anchored Fusion of Multi-Ratio DiT Reconstruction Residuals for Cross-Domain Audio Deepfake Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026ToolDF: Tool-Integrated Reasoning for Mixed-Authenticity Audio Deepfake Detection
  2. 2026Leveraging large multimodal models for audio-video deepfake detection: a pilot study
  3. 2024Diffuse or Confuse: A Diffusion Deepfake Speech Dataset
  4. 2026Interpretable All-Type Audio Deepfake Detection with Audio LLMs via Frequency-Time Reinforcement Learning
  5. 2025AUDETER: A Large-scale Dataset for Deepfake Audio Detection in Open Worlds
  6. 2024RawBMamba: End-to-End Bidirectional State Space Model for Audio Deepfake Detection
  7. 2026MADBench: A Benchmark for Modality-Aware Audio Deepfake Detection
  8. 2026Leveraging Gradient Reversal Loss and Multitask Learning for Datasets-Aware Audio Deepfake Detection