AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

VoiceBlock: Privacy through Real-Time Adversarial Attacks with Audio-to-Audio Models

Patrick O'Reilly, Andreas Bugler, Keshav Bhandari, Max Morrison, Bryan Pardo

NeurIPS 2022 · Adversarial robustness and security › Attacks on text and graphs · approach: Diffusion-based adversarial purification · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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 changedWhereMeasuredEffect
Use VoiceBlock instead of universal additive perturbationspeaker de-identification · VoxCeleb1 qualityPESQ: 1.35 -> 3.74Improved
Use VoiceBlock instead of universal additive perturbationspeaker de-identification · VoxCeleb1 qualitySTOI: 0.78 -> 0.92Improved
Use VoiceBlock instead of spectral gatingspeaker de-identification · VoxCeleb1 qualityPESQ: 1.11 -> 3.74Improved
Use VoiceBlock instead of white noisespeaker de-identification · VoxCeleb1 qualityPESQ: 1.03 -> 3.74Improved
Use VoiceBlock instead of universal perturbationResNetSE34v2 (seen) · VoxCeleb1top-1 accuracy: 0.09 -> 0.03Mixed
Use VoiceBlock instead of spectral gatingResNetSE34v2 (seen) · VoxCeleb1top-1 accuracy: 0.02 -> 0.03No clear effect
Use VoiceBlock instead of universal perturbationY-Vector (unseen) · VoxCeleb1top-1 accuracy: 0.48 -> 0.35Improved
Use spectral gating instead of VoiceBlockY-Vector (unseen) · VoxCeleb1top-1 accuracy: 0.35 -> 0.03Improved
Use white noise instead of VoiceBlockY-Vector (unseen) · VoxCeleb1top-1 accuracy: 0.35 -> 0.00Improved
Apply VoiceBlock vs no attackResNetSE34v2 · VoxCeleb1top-1 accuracy: 0.99 -> 0.03Improved
Apply VoiceBlock vs no attackY-Vector (unseen) · VoxCeleb1top-1 accuracy: 0.96 -> 0.35Improved
Use VoiceBlock instead of other attacks in listening testMUSHRA listening study · 100 VoxCeleb1 clipsquality rating significance: p = 0.024Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy