PulseAugur
EN
LIVE 15:07:23

New research details backdoor attacks and defenses for speech recognition models

By PulseAugur Editorial · [3 sources] ·

Two new research papers explore the vulnerabilities of speech recognition models to backdoor attacks. The first paper introduces SpeechGuard, a system designed to detect and neutralize these attacks in real-time by identifying and purifying poisoned audio samples. The second paper details "natural backdoor attacks," which use everyday sounds as triggers, demonstrating high success rates even with subtle or short triggers and posing a significant, difficult-to-detect threat. AI

IMPACT Highlights critical security vulnerabilities in speech recognition systems, potentially impacting applications in sensitive areas like autonomous driving.

RANK_REASON Two academic papers published on arXiv detailing new research into backdoor attacks on speech recognition models and a proposed defense mechanism.

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

New research details backdoor attacks and defenses for speech recognition models

COVERAGE [3]

  1. arXiv cs.AI TIER_1 English(EN) · Ye Lu, Yihan Yan, Zhaoyang Zhang, Zhitao Ou, Runze Liu, Li Liu, Shen Wang ·

    Do Speech Tokens Leak Voiceprints? Speaker Inversion Attacks Against End-to-End Speech Language Models

    arXiv:2607.16870v1 Announce Type: cross Abstract: End-to-end speech language models increasingly represent user speech with speech tokens rather than relying exclusively on cascaded ASR--LLM--TTS pipelines. Although these tokens support expressive and low-latency spoken interacti…

  2. arXiv cs.LG TIER_1 English(EN) · Jinwen Xin, Xixiang Lv ·

    SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models

    arXiv:2607.15697v1 Announce Type: cross Abstract: Backdoor attacks pose a critical threat to neural network models, allowing attackers to implant a backdoor during the training phase by manipulating a small portion of the training data. In security-sensitive applications such as …

  3. arXiv cs.LG TIER_1 English(EN) · Jinwen Xin, Xixiang Lyu, Jing Ma ·

    Natural Backdoor Attacks on Speech Recognition Models

    arXiv:2607.15724v1 Announce Type: cross Abstract: With the rapid development of deep learning, its vulnerability has gradually emerged in recent years. This work focuses on backdoor attacks on speech recognition systems. We adopt sounds that are ordinary in nature or in our daily…

RELATED ENTITIES

RELATED TOPICS