Defend Against Adversarial Samples by Using Perceptual Hash

Liu, Changrui; Ye, Dengpan; Shang, Yongfeng; Jiang, Shunzhi; Li, Shiyu; Mei, Yuan; Wang, Liqiang

doi:10.32604/cmc.2020.07421

Cited by 8 publications

(8 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the meantime, many efforts have been devoted to defend against adversarial examples 32–37 . Existing research try to implement defense in three different categories: robust optimization, certified robustness, and detection defense.…”

Section: Adversarial Defense and Detectionmentioning

confidence: 99%

“…In the meantime, many efforts have been devoted to defend against adversarial examples. [32][33][34][35][36][37] Existing research try to implement defense in three different categories: robust optimization, certified robustness, and detection defense. In the robust optimization, various adversarial training methods [27][28][29] belong to the kind of effective defense technologies against adversarial attacks.…”

Section: Adversarial Defense and Detectionmentioning

confidence: 99%

See 1 more Smart Citation

Detection defense against adversarial attacks with saliency map

Chen

Liu

et al. 2021

Int J of Intelligent Sys

Self Cite

View full text Add to dashboard Cite

It is well established that neural networks are vulnerable to adversarial examples, which are almost imperceptible on human vision and can cause the deep models misbehave. Such phenomenon may lead to severely inestimable consequences in the safety and security critical applications. Existing defenses are trend to harden the robustness of models against adversarial attacks, for example, adversarial training technology. However, these are usually intractable to implement due to the high cost of retraining and the cumbersome operations of altering the model architecture or parameters. In this paper, we discuss the saliency map method from the view of enhancing model interpretability, it is similar to introducing the mechanism of the attention to the model, so as to comprehend the progress of object identification by the deep networks. We then propose a novel method combined with additional noises and utilize the inconsistency strategy to detect adversarial examples. Our experimental results of some representative adversarial attacks on common data sets including ImageNet and popular models show that our method can detect all the attacks with high detection success rate effectively. We compare it with the existing state‐of‐the‐art technique, and the experiments indicate that our method is more general.

show abstract

Section: Adversarial Defense and Detectionmentioning

confidence: 99%

Section: Adversarial Defense and Detectionmentioning

confidence: 99%

Detection defense against adversarial attacks with saliency map

Chen

Liu

et al. 2021

Int J of Intelligent Sys

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the meantime, many efforts have been devoted to defend against adversarial examples (Xie et al 2018;Guo et al 2018;Samangouei et al 2018;Song et al 2017;Liao et al 2018;Liu et al 2020;Yang et al 2020;Wang et al 2020). Existing researches try to implement defense in three different categories: robust optimization, certified robustness and detection defense.…”

Section: B Adversarial Defense and Detectionmentioning

confidence: 99%

Detection Defense Against Adversarial Attacks with Saliency Map

Chen

Liu

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

It is well established that neural networks are vulnerable to adversarial examples, which are almost imperceptible on human vision and can cause the deep models misbehave. Such phenomenon may lead to severely inestimable consequences in the safety and security critical applications. Existing defenses are trend to harden the robustness of models against adversarial attacks, e.g., adversarial training technology. However, these are usually intractable to implement due to the high cost of retraining and the cumbersome operations of altering the model architecture or parameters. In this paper, we discuss the saliency map method from the view of enhancing model interpretability, it is similar to introducing the mechanism of the attention to the model, so as to comprehend the progress of object identification by the deep networks. We then propose a novel method combined with additional noises and utilize the inconsistency strategy to detect adversarial examples. Our experimental results of some representative adversarial attacks on common datasets including ImageNet and popular models show that our method can detect all the attacks with high detection success rate effectively. We compare it with the existing state-of-the-art technique, and the experiments indicate that our method is more general.

show abstract

“…As shown in Fig. 2, for one of the videos used in this study, most of the frames of the NIR video are near duplicates when compared using perceptual hashing [15,19]. Attempting to augment the training set by, for example, (i) tiling video's initial frame with many ROIs (see Fig.…”

Section: Related Workmentioning

confidence: 99%

“…Fig.2. Perceptual hashing algorithm[15,19] applied on one of the perfusion videos used in this study. Selected images till the end of the perfusion t = 860s are near duplicates of the first frame.…”

mentioning

confidence: 99%

Perfusion Quantification from Endoscopic Videos: Learning to Read Tumor Signatures

Zhuk¹,

Epperlein²,

Nair³

et al. 2020

Preprint

View full text Add to dashboard Cite

Intra-operative identification of malignant versus benign or healthy tissue is a major challenge in fluorescence guided cancer surgery. We propose a perfusion quantification method for computer-aided interpretation of subtle differences in dynamic perfusion patterns which can be used to distinguish between normal tissue and benign or malignant tumors intra-operatively in real-time by using multispectral endoscopic videos. The method exploits the fact that vasculature arising from cancer angiogenesis gives tumors differing perfusion patterns from the surrounding tissue, and defines a signature of tumor which could be used to differentiate tumors from normal tissues. Experimental evaluation of our method on a cohort of colorectal cancer surgery endoscopic videos suggests that the proposed tumor signature is able to successfully discriminate between healthy, cancerous and benign tissue with 95% accuracy.

show abstract

Defend Against Adversarial Samples by Using Perceptual Hash

Cited by 8 publications

References 8 publications

Detection defense against adversarial attacks with saliency map

Detection defense against adversarial attacks with saliency map

Detection Defense Against Adversarial Attacks with Saliency Map

Perfusion Quantification from Endoscopic Videos: Learning to Read Tumor Signatures

Contact Info

Product

Resources

About