•  
  •  
 

Abstract

The emergence of "Deep Learning" DL has revolutionized the scope of cybersecurity and computer vision. However, this technology is not immune to emerging challenges that can affect its performance and security. One major challenge is the availability of large datasets for training DL algorithms. Furthermore, there is a need for improved algorithms and architectures that can effectively process such datasets. Another challenge is the constant evolution of cyber threats, which require the development of new DL models to defend against them. Additionally, the interpretability and explain ability of DL models in cybersecurity pose a significant challenge, as their black-box nature can make them difficult to understand and mitigate against. Therefore, the emerging challenges in DL for computer vision and cybersecurity require a coordinated effort from researchers and practitioners in the felid of neural network specially with generative adversarial network to overcome handicaps and effectively leverage the technology to enhance security and surveillance in various domains.

Keywords

Adversarial attacks, Deep learning (DL), Generative adversarial network (GAN), Deep neural network (DNN), Computer vision, Cybersecurity

Subject Area

Computer Science

Article Type

Article

First Page

2791

Last Page

2799

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Download

Share

COinS