SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models

arXiv:2607.15697 · cs.SD, cs.CR, cs.LG · Submitted 2026-07-17 · Read on arXiv

Jinwen Xin, Xixiang Lv

cs.SD, cs.CR, cs.LG

Submitted: 2026-07-17

Comments: 8 pages

Journal ref: 2024 International Joint Conference on Neural Networks (IJCNN), pp. 1-8, 2024

DOI: 10.1109/IJCNN60899.2024.10650300

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Accuracy — 19× in this paper · explained in 1 episode
• Attack Success Rate — 3× in this paper
• Activation Function — 2× in this paper · explained in 1 episode
• Verification — 2× in this paper · explained in 2 episodes
• Attack Success Rate (ASR) — 1× in this paper · explained in 2 episodes
• Attention Mechanism — 1× in this paper · explained in 3 episodes
• Benign Accuracy — 1× in this paper
• Clustering — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Decode — 1× in this paper · explained in 2 episodes
• False Acceptance Rate — 1× in this paper
• False Rejection Rate — 1× in this paper
• In Network and Distributed Systems Security — 1× in this paper
• Loss Function — 1× in this paper · explained in 2 episodes
• Pruning — 1× in this paper · explained in 4 episodes
• Purification Accuracy — 1× in this paper
• Regularization — 1× in this paper · explained in 3 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Security and Privacy — 1× in this paper
• Short-Time Fourier Transform — 1× in this paper
• Sparsity — 1× in this paper · explained in 4 episodes
• Transfer Learning — 1× in this paper · explained in 10 episodes

Sources

• Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning
• Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition
• AudioMNIST: Exploring Explainable Artificial Intelligence for Audio Analysis on a Simple Benchmark
• A neural attention model for speech command recognition

Related papers

• Few-Shot Open-Set Audio Classification via Transductive Prototype Refinement and Class Logit Enhancement
• Spectral Masking and Interpolation Attack (SMIA): A Black-box Adversarial Attack against Voice Authentication and Anti-Spoofing Systems
• AVMeme Exam: A Multimodal Multilingual Multicultural Benchmark for LLMs' Contextual and Cultural Knowledge and Thinking
• SoundWeaver: Compositional Warm-Starting for Text-to-Audio Diffusion Serving
• WASIL: In-the-Wild Arabic Spoken Interactions with LLMs
• Efficient Audiovisual Speech Processing via MUTUD: Multimodal Training and Unimodal Deployment