Universal Adversarial Perturbations For Speech Recognition Systems

Abstract

In this work, we demonstrate the existence of universal adversarial audio perturbations that cause mis-transcription of audio signals by automatic speech recognition (ASR) systems. We propose an algorithm to find a single quasi-imperceptible perturbation, which when added to any arbitrary speech signal, will most likely fool the victim speech recognition model. Our experiments demonstrate the application of our proposed technique by crafting audio-agnostic universal perturbations for the state-of-the-art ASR system -- Mozilla DeepSpeech. Additionally, we show that such perturbations generalize to a significant extent across models that are not available during training, by performing a transferability test on a WaveNet based ASR system.

Stats

• citations75
• S2 citations—
• github stars0
• HF likes0
• heat score14.11
• arxiv keyneekhara2019universal

Related papers

• Universal Adversarial Examples In Speech Command Classification (2019)0.00
• Universal Adversarial Perturbations Generative Network For Speaker Recognition (2020)12.33
• Exploring Targeted Universal Adversarial Perturbations To End-to-end ASR Models (2021)8.35
• Inaudible Adversarial Perturbations For Targeted Attack In Speaker Recognition (2020)12.33
• Targeted Adversarial Examples For Black Box Audio Systems (2018)15.75
• Impact Of Phonetics On Speaker Identity In Adversarial Voice Attack (2025)0.00
• Asynchronous Voice Anonymization Using Adversarial Perturbation On Speaker Embedding (2024)7.16
• Adversarial Attacks Against Automatic Speech Recognition Systems Via Psychoacoustic Hiding (2018)16.45