Deep Neural Networks for Pattern Recognition

TL;DR

This paper discusses the structure, functions, and recent training strategies of deep neural networks in pattern recognition, highlighting their human-like accuracy in tasks like image classification and object detection.

Contribution

It provides an overview of deep neural network architectures, their relation to human visual perception, and recent advancements in training methods for complex models.

Findings

Deep neural networks achieve human-level accuracy in image tasks

Conditional generative adversarial networks mimic visual perception processes

Recent training strategies improve learning efficiency of complex networks

Abstract

In the field of pattern recognition research, the method of using deep neural networks based on improved computing hardware recently attracted attention because of their superior accuracy compared to conventional methods. Deep neural networks simulate the human visual system and achieve human equivalent accuracy in image classification, object detection, and segmentation. This chapter introduces the basic structure of deep neural networks that simulate human neural networks. Then we identify the operational processes and applications of conditional generative adversarial networks, which are being actively researched based on the bottom-up and top-down mechanisms, the most important functions of the human visual perception process. Finally, recent developments in training strategies for effective learning of complex deep neural networks are addressed.