Role of Deep Learning in Adversarial Attacks: A Survey
Priya Sachdeva, Raman Kumar · 2024
Deep learning has become an effective paradigm that is impacting several fields, including decision-making systems, natural language processing, and recognition of images. Nonetheless, the growing frequency of malicious attacks presents an important barrier to the robustness and dependability of deep learning models. This research study explores how adversarial attacks use deep learning and the techniques by which malicious actors take advantage of neural network weaknesses. An overview of a variety of adversarial attacks and the most recent methods for detecting and preventing them on diverse data sets are given in this study.