A Systematic Evaluation of Adversarial Attacks against Speech Emotion Recognition Models

TL;DR

This paper systematically evaluates the vulnerability of CNN-LSTM speech emotion recognition models to various adversarial attacks across different languages and genders, highlighting significant robustness issues and providing insights for future improvements.

Contribution

It introduces a methodology for assessing adversarial robustness in SER models and provides comprehensive analysis of attack impacts across languages and genders, serving as a baseline for future defenses.

Findings

Adversarial attacks significantly reduce SER model performance.

CNN-LSTM models are highly vulnerable to both white-box and black-box attacks.

Minor differences in attack efficacy across languages and genders.

Abstract

Speech emotion recognition (SER) is constantly gaining attention in recent years due to its potential applications in diverse fields and thanks to the possibility offered by deep learning technologies. However, recent studies have shown that deep learning models can be vulnerable to adversarial attacks. In this paper, we systematically assess this problem by examining the impact of various adversarial white-box and black-box attacks on different languages and genders within the context of SER. We first propose a suitable methodology for audio data processing, feature extraction, and CNN-LSTM architecture. The observed outcomes highlighted the significant vulnerability of CNN-LSTM models to adversarial examples (AEs). In fact, all the considered adversarial attacks are able to significantly reduce the performance of the constructed models. Furthermore, when assessing the efficacy of the attacks, minor differences were noted between the languages analyzed as well as between male and female speech. In summary, this work contributes to the understanding of the robustness of CNN-LSTM models, particularly in SER scenarios, and the impact of AEs. Interestingly, our findings serve as a baseline for a) developing more robust algorithms for SER, b) designing more effective attacks, c) investigating possible defenses, d) improved understanding of the vocal differences between different languages and genders, and e) overall, enhancing our comprehension of the SER task.