A Robust Approach for Securing Audio Classification Against Adversarial Attacks

Abstract

Adversarial audio attacks can be considered as a small perturbation unperceptive to human ears that is intentionally added to an audio signal and causes a machine learning model to make mistakes. This poses a security concern about the safety of machine learning models since the adversarial attacks can fool such models toward the wrong predictions. In this paper we first review some strong adversarial attacks that may affect both audio signals and their 2D representations and evaluate the resiliency of deep learning models and support vector machines (SVM) trained on 2D audio representations such as short time Fourier transform, discrete wavelet transform (DWT) and cross recurrent plot against several state-of-the-art adversarial attacks. Next, we propose a novel approach based on pre-processed DWT representation of audio signals and SVM to secure audio systems against adversarial attacks. The proposed architecture has several preprocessing modules for generating and enhancing spectrograms including dimension reduction and smoothing. We extract features from small patches of the spectrograms using the speeded up robust feature (SURF) algorithm which are further used to transform into cluster distance distribution using the K-Means++ algorithm. Finally, SURF-generated vectors are encoded by this codebook and the resulting codewords are used for training a SVM. All these steps yield to a novel approach for audio classification that provides a good tradeoff between accuracy and resilience. Experimental results on three environmental sound datasets show the competitive performance of the proposed approach compared to the deep neural networks both in terms of accuracy and robustness against strong adversarial attacks.

Detection of Adversarial Attacks and Characterization of Adversarial Subspace

“…Thus far, the best sound classification accuracy has been achieved for deep learning algorithms trained on 2D signal representations [1,2]. However, it has been shown that despite achieving high performance, the approaches based on 2D representations are very vulnerable against adversarial attacks [3]. Unfortunately, this poses a strict security issue because crafted adversarial examples not only mislead the target model toward a wrong label, but also, they are transferable to other models including conventional algorithms such as support vector machines (SVM) [3].…”

A Fast Online Cascaded Regression Algorithm for Face Alignment

“…A cascade regression of extreme learning machine is first introduced and its parallel version is developed to train a offline model. Then, we present an efficient method to incrementally update a trained model to make it more generalizable [16], [43] and new optimization strategy [44], [45].…”

Unsupervised feature learning for environmental sound classification using Weighted Cycle-Consistent Generative Adversarial Network

“…The motivation behind proposing a shallow approach instead of a deep architecture as a front-end classifier is twofold. First, it has been shown that advanced deep neural networks such as AlexNet, GoogLeNet and other recent architectures are highly vulnerable to adversarial attacks as they can predict wrong labels with high confidence [60,61]. Secondly, conventional classifiers such as SVMs and RFs, which learn from handcrafted features are considerably more robust against such adversarial attacks than deep learning models [60].…”

“…First, it has been shown that advanced deep neural networks such as AlexNet, GoogLeNet and other recent architectures are highly vulnerable to adversarial attacks as they can predict wrong labels with high confidence [60,61]. Secondly, conventional classifiers such as SVMs and RFs, which learn from handcrafted features are considerably more robust against such adversarial attacks than deep learning models [60]. Taking advantage of these two facts, we propose a conventional data-driven model as front-end classifier and use a generative model based on a deep architecture as a back-end classifier for data augmentation purposes only.…”