English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep learning models have been used in creating various effective imageclassification applications. However, they are vulnerable to adversarialattacks that seek to misguide the models into predicting incorrect classes. Ourstudy of major adversarial attack models shows that they all specificallytarget and exploit the neural networking structures in their designs. Thisunderstanding makes us develop a hypothesis that most classical machinelearning models, such as Random Forest (RF), are immune to adversarial attackmodels because they do not rely on neural network design at all. Ourexperimental study of classical machine learning models against popularadversarial attacks supports this hypothesis. Based on this hypothesis, wepropose a new adversarial-aware deep learning system by using a classicalmachine learning model as the secondary verification system to complement theprimary deep learning model in image classification. Although the secondaryclassical machine learning model has less accurate output, it is only used forverification purposes, which does not impact the output accuracy of the primarydeep learning model, and at the same time, can effectively detect anadversarial attack when a clear mismatch occurs. Our experiments based onCIFAR-100 dataset show that our proposed approach outperforms currentstate-of-the-art adversarial defense systems.

Adversarial-Aware Deep Learning System based on a Secondary Classical Machine Learning Verification Approach