EmoAttack: Utilizing Emotional Voice Conversion for Speech Backdoor Attacks on Deep Speech Classification Models
On the map: LLM safety and security → Audio jailbreaks and speech backdoors
Abstract (arXiv)
Deep speech classification tasks, mainly including keyword spotting and speaker verification, play a crucial role in speech-based human-computer interaction. Recently, the security of these technologies has been demonstrated to be vulnerable to backdoor attacks. Specifically speaking, speech samples are attacked by noisy disruption and component modification in present triggers. We suggest that speech backdoor attacks can strategically focus on emotion, a higher-level subjective perceptual attribute inherent in speech. Furthermore, we proposed that emotional voice conversion technology can serve as the speech backdoor attack trigger, and the method is called EmoAttack. Based on this, we conducted attack experiments on two speech classification tasks, showcasing that EmoAttack method owns impactful trigger effectiveness and its remarkable attack success rate and accuracy variance. Additionally, the ablation experiments found that speech with intensive emotion is more suitable to be targeted for attacks.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in Audio jailbreaks and speech backdoors, one of 12 directions in LLM safety and security. Builds on 24 of its 33 references that are in the map. First posted on arXiv 28 Aug 2024.
Builds on
24 of 33 references.
Closest ancestors · 4
Same direction, newest first
- 2023Towards Stealthy Backdoor Attacks against Speech Recognition via Elements of Sound17 Jul
- 2023Fake the Real: Backdoor Attack on Deep Speech Classification via Voice Conversion28 Jun
- 2022VSVC: Backdoor attack against Keyword Spotting based on Voiceprint Selection and Voice Conversion20 Dec
- 2022PBSM: Backdoor attack against Keyword spotting based on pitch boosting and sound masking16 Nov
Other references · 20
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
- 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,052
- 2018Speech Commands: A Dataset for Limited-Vocabulary Speech Recognitioncited by 741
- 2017VoxCeleb: a large-scale speaker identification datasetcited by 714
- 2020Backdoor Learning: A Surveycited by 412
- 2017Generalized End-to-End Loss for Speaker Verificationcited by 290
- 2017A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Predictioncited by 245
- 2019A backdoor attack against LSTM-based text classification systemscited by 214
- 2020Blind Backdoors in Deep Learning Modelscited by 174
- 2018Speaker Recognition from Raw Waveform with SincNetcited by 166
- 2021Emotional Voice Conversion: Theory, Databases and ESDcited by 113
8 more
- 2020Mitigating backdoor attacks in LSTM-based Text Classification Systems by Backdoor Keyword Identificationcited by 80
- 2020Backdoor Attack against Speaker Verificationcited by 62
- 2021Keyword Transformer: A Self-Attention Model for Keyword Spottingcited by 59
- 2021Can You Hear It? Backdoor Attacks via Ultrasonic Triggerscited by 41
- 2022Going In Style: Audio Backdoors Through Stylistic Transformationscited by 19
- 2020Natural Backdoor Attack on Text Datacited by 18
- 2021An Improved StarGAN for Emotional Voice Conversion: Enhancing Voice Quality and Data Augmentationcited by 7
- 2024SER Evals: In-domain and Out-of-domain Benchmarking for Speech Emotion Recognitioncited by 6
Built on by
Citing papers per year: 2025: 1; 2026: 1.
2 papers.