Secure machine learning against adversarial samples at test time

Deep neural networks (DNNs) are widely used to handle many difficult tasks, such as image classification and malware detection, and achieve outstanding performance. However, recent studies on adversarial examples, which have maliciously undetectable perturbations added to their original samples that are indistinguishable by human eyes but mislead the machine learning approaches, show that machine learning models are vulnerable to security attacks. Though various adversarial retraining techniques have been developed in the past few years, none of them is scalable. In this paper, we propose a new iterative adversarial retraining approach to robustify the model and to reduce the effectiveness of adversarial inputs on DNN models. The proposed method retrains the model with both Gaussian noise augmentation and adversarial generation techniques for better generalization. Furthermore, the ensemble model is utilized during the testing phase in order to increase the robust test accuracy. The results from our extensive experiments demonstrate that the proposed approach increases the robustness of the DNN model against various adversarial attacks, specifically, fast gradient sign attack, Carlini and Wagner (C&W) attack, Projected Gradient Descent (PGD) attack, and DeepFool attack. To be precise, the robust classifier obtained by our proposed approach can maintain a performance accuracy of 99% on average on the standard test set. Moreover, we empirically evaluate the runtime of two of the most effective adversarial attacks, i.e., C&W attack and BIM attack, to find that the C&W attack can utilize GPU for faster adversarial example generation than the BIM attack can. For this reason, we further develop a parallel implementation of the proposed approach. This parallel implementation makes the proposed approach scalable for large datasets and complex models.

Adversarial example defense based on image reconstruction

The rapid development of deep neural networks (DNN) has promoted the widespread application of image recognition, natural language processing, and autonomous driving. However, DNN is vulnerable to adversarial examples, such as an input sample with imperceptible perturbation which can easily invalidate the DNN and even deliberately modify the classification results. Therefore, this article proposes a preprocessing defense framework based on image compression reconstruction to achieve adversarial example defense. Firstly, the defense framework performs pixel depth compression on the input image based on the sensitivity of the adversarial example to eliminate adversarial perturbations. Secondly, we use the super-resolution image reconstruction network to restore the image quality and then map the adversarial example to the clean image. Therefore, there is no need to modify the network structure of the classifier model, and it can be easily combined with other defense methods. Finally, we evaluate the algorithm with MNIST, Fashion-MNIST, and CIFAR-10 datasets; the experimental results show that our approach outperforms current techniques in the task of defending against adversarial example attacks.

Semantic-Preserving Adversarial Text Attacks

Deep learning models are known immensely brittle to adversarial image examples, yet their vulnerability in text classification is insufficiently explored. Existing text adversarial attack strategies can be roughly divided into three categories, i.e., character-level attack, word-level attack, and sentence-level attack. Despite the success brought by recent text attack methods, how to induce misclassification with the minimal text modifications while keeping the lexical correctness, syntactic soundness, and semantic consistency simultaneously is still a challenge. To examine the vulnerability of deep models, we devise a Bigram and Unigram based adaptive Semantic Preservation Optimization (BU-SPO) approach which attacks text documents not only at a unigram word level but also at a bigram level to avoid generating meaningless sentences. We also present a hybrid attack strategy that collects substitution words from both synonyms and sememe candidates, to enrich the potential candidate set. Besides, a Semantic Preservation Optimization (SPO) method is devised to determine the word substitution priority and reduce the perturbation cost. Furthermore, we constraint the SPO with a semantic Filter (dubbed SPOF) to improve the semantic similarity between the input text and the adversarial example. To estimate the effectiveness of our proposed methods, BU-SPO and BU-SPOF, we attack four victim deep learning models trained on three real-world text datasets. Experimental results demonstrate that our approaches accomplish the highest semantics consistency and attack success rates by making the minimal word modifications compared with competitive methods.

Semantic-Preserving Adversarial Text Attacks

Deep learning models are known immensely brittle to adversarial image examples, yet their vulnerability in text classification is insufficiently explored. Existing text adversarial attack strategies can be roughly divided into three categories, i.e., character-level attack, word-level attack, and sentence-level attack. Despite the success brought by recent text attack methods, how to induce misclassification with the minimal text modifications while keeping the lexical correctness, syntactic soundness, and semantic consistency simultaneously is still a challenge. To examine the vulnerability of deep models, we devise a Bigram and Unigram based adaptive Semantic Preservation Optimization (BU-SPO) approach which attacks text documents not only at a unigram word level but also at a bigram level to avoid generating meaningless sentences. We also present a hybrid attack strategy that collects substitution words from both synonyms and sememe candidates, to enrich the potential candidate set. Besides, a Semantic Preservation Optimization (SPO) method is devised to determine the word substitution priority and reduce the perturbation cost. Furthermore, we constraint the SPO with a semantic Filter (dubbed SPOF) to improve the semantic similarity between the input text and the adversarial example. To estimate the effectiveness of our proposed methods, BU-SPO and BU-SPOF, we attack four victim deep learning models trained on three real-world text datasets. Experimental results demonstrate that our approaches accomplish the highest semantics consistency and attack success rates by making the minimal word modifications compared with competitive methods.

Opportunistic Use of Crowdsourced Workers for Online Relabeling of Potential Adversarial Examples

<div>Deep Neural Networks (DDNs) have achieved tremendous success in handling various Machine Learning (ML) tasks, such as speech recognition, Natural Language Processing, and image classification. However, they have shown vulnerability to well-designed inputs called adversarial examples. Researchers in industry and academia have proposed many adversarial example defense techniques. However, none can provide complete robustness. The cutting-edge defense techniques offer partial reliability. Thus, complementing them with another layer of protection is a must, especially for mission-critical applications. This paper proposes a novel Online Selection and Relabeling Algorithm (OSRA) that opportunistically utilizes a limited number of crowdsourced workers (budget-constraint crowdsourcing) to maximize the ML system’s robustness. OSRA strives to use crowdsourced workers effectively by selecting the most suspicious inputs (the potential adversarial examples) and moving them to the crowdsourced workers to be validated and corrected (relabeled). As a result, the impact of adversarial examples gets reduced, and accordingly, the ML system becomes more robust. We also proposed a heuristic threshold selection method that contributes to enhancing the prediction system’s reliability. We empirically validated our proposed algorithm and found that it can efficiently and optimally utilize the allocated budget for crowdsourcing. It is also effectively integrated with a state-ofthe- art black-box (transfer-based) defense technique, resulting in a more robust system. Simulation results show that OSRA can outperform a random selection algorithm by 60% and achieve comparable performance to an optimal offline selection benchmark. They also show that OSRA’s performance has a positive correlation with system robustness.<br></div>

Opportunistic Use of Crowdsourced Workers for Online Relabeling of Potential Adversarial Examples

<div>Deep Neural Networks (DDNs) have achieved tremendous success in handling various Machine Learning (ML) tasks, such as speech recognition, Natural Language Processing, and image classification. However, they have shown vulnerability to well-designed inputs called adversarial examples. Researchers in industry and academia have proposed many adversarial example defense techniques. However, none can provide complete robustness. The cutting-edge defense techniques offer partial reliability. Thus, complementing them with another layer of protection is a must, especially for mission-critical applications. This paper proposes a novel Online Selection and Relabeling Algorithm (OSRA) that opportunistically utilizes a limited number of crowdsourced workers (budget-constraint crowdsourcing) to maximize the ML system’s robustness. OSRA strives to use crowdsourced workers effectively by selecting the most suspicious inputs (the potential adversarial examples) and moving them to the crowdsourced workers to be validated and corrected (relabeled). As a result, the impact of adversarial examples gets reduced, and accordingly, the ML system becomes more robust. We also proposed a heuristic threshold selection method that contributes to enhancing the prediction system’s reliability. We empirically validated our proposed algorithm and found that it can efficiently and optimally utilize the allocated budget for crowdsourcing. It is also effectively integrated with a state-ofthe- art black-box (transfer-based) defense technique, resulting in a more robust system. Simulation results show that OSRA can outperform a random selection algorithm by 60% and achieve comparable performance to an optimal offline selection benchmark. They also show that OSRA’s performance has a positive correlation with system robustness.<br></div>