Enhancing Image Classification Performance Through Deep Residual Learning Networks

EasyChair Preprint 15811

Abstract

In recent years, deep neural networks have achieved significant breakthroughs in image recognition tasks. One of the main challenges in this domain is the degradation of model performance as the network depth increases. In this paper, we explore Residual Neural Networks (ResNet), which allow for the construction of very deep models without performance degradation by utilizing shortcut connections between layers. Our experimental results show that employing residual architectures, particularly in deeper networks, can substantially improve image recognition performance. This research could have widespread applications in deep learning projects related to image recognition and computer vision.

Keyphrases: Applications, Deep Neural Network, ResNet, deep learning

@booklet{EasyChair:15811,
  author    = {Amir Arslan and Xi Zhang and Ahmad Hossein},
  title     = {Enhancing Image Classification Performance Through Deep Residual Learning Networks},
  howpublished = {EasyChair Preprint 15811},
  year      = {EasyChair, 2025}}