Research map / Adversarial robustness and security / Attacks on text and graphs
Defending against Adversarial Audio via Diffusion Model
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add waveform diffusion purification before classification | SC09 acoustic classifier · L∞ white-box PGD100 | robust accuracy: 71 -> 84 | Helped |
| Add waveform diffusion purification before classification | SC09 acoustic classifier · L2 white-box PGD100 | robust accuracy: 65 -> 84 | Helped |
| Add waveform diffusion purification before classification | SC09 acoustic classifier · L∞ black-box FAKEBOB | robust accuracy: 92 -> 86 | Hurt |
| Replace waveform purification with Mel-spectrogram purification (ablation) | AudioPure · SC09 L∞ white-box PGD70 | robust accuracy: 84 -> 72 | Hurt |
| Replace waveform purification with Mel-spectrogram purification (ablation) | AudioPure · SC09 L2 white-box PGD100 | robust accuracy: 84 -> 49 | Hurt |
| Increase diffusion steps from n*=1 to n*=3 (ablation) | AudioPure · SC09 L∞ PGD10, ε=0.002 | robust accuracy: 94 -> 89 | Hurt |
| Increase diffusion steps from n*=1 to n*=3 (ablation) | AudioPure · SC09 L∞ PGD10, ε=0.008 | robust accuracy: 27 -> 85 | Helped |
| Fine-tune DiffWave on SSP adversarial examples (ablation) | AudioPure · SC09 L∞ PGD70 | robust accuracy: 84 -> 82 | Hurt |
| Fine-tune DiffWave on SSP adversarial examples (ablation) | AudioPure · SC09 L∞ PGD10 | robust accuracy: 89 -> 89 | No clear effect |
| Use AudioPure instead of no defense | Qualcomm Keyword Speech Dataset classifier · L∞ PGD10, ε=0.002 | robust accuracy: 0 -> 61 | Helped |
| Increase attack iterations (ablation) | Adaptive PGD against AudioPure · SC09 | robust accuracy | No clear effect |
| Increase EOT sample size (ablation) | Adaptive EOT attack against AudioPure · SC09 | robust accuracy | No clear effect |
Closest papers by meaning
- AdvWave: Stealthy Adversarial Jailbreak Attack against Large Audio-Language Models (ICLR 2025 · similarity 0.88)
- Rethinking Audio-Visual Adversarial Vulnerability from Temporal and Modality Perspectives (ICLR 2025 · similarity 0.86)
- Divide and Conquer: Heterogeneous Noise Integration for Diffusion-based Adversarial Purification (CVPR 2025 · similarity 0.86)
- I Can Hear You: Selective Robust Training for Deepfake Audio Detection (ICLR 2025 · similarity 0.85)
- Adaptive Compression and Targeted Perturbation: A Unified Framework for Generalized Audio Deepfake Detection (NeurIPS 2026 · similarity 0.85)
- ADBM: Adversarial Diffusion Bridge Model for Reliable Adversarial Purification (ICLR 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.