An Introduction to Deep Convolutional Neural Networks With Keras

An Introduction to Deep Convolutional Neural Networks With Keras

Wazir Muhammad (Electrical Engineering Department, BUET, Khuzdar, Pakistan), Irfan Ullah (Department of Electrical Engineering, Chulalongkorn University, Bangkok, Thailand) and Mohammad Ashfaq (School of Life Sciences, B. S. Abdur Rahman Crescent Institute of Science and Technology, Chennai, India)
Copyright: © 2020 |Pages: 42
DOI: 10.4018/978-1-7998-3095-5.ch011

Abstract

Deep learning (DL) is the new buzzword for researchers in the research area of computer vision that unlocked the doors to solving complex problems. With the assistance of Keras library, machine learning (ML)-based DL and various complicated or unresolved issues such as face recognition and voice recognition might be resolved easily. This chapter focuses on the basic concept of Keras-based framework DL library to handle the different real-life problems. The authors discuss the codes of previous libraries and same code run on Keras library and assess the performance on Google Colab Cloud Graphics Processing Units (GPUs). The goal of this chapter is to provide you with the newer concept, algorithm, and technology to solve the real-life problems with the help of Keras framework. Moreover, they discuss how to write the code of standard convolutional neural network (CNN) architectures using Keras libraries. Finally, the codes of validation and training data set to start the training procedure are explored.
Chapter Preview
Top

Introduction

Artificial Intelligence (AI) or machine intelligence, is mainly the simulation of the natural intelligence of humans with the help of machines. AI systems are able to learn and recognized the configurations to reach any decisions as well as conclusions on the basis of different analytical situations, thereby the utmost chance of successfully accomplishing the goals (Dreyfus, 1979). Usually, machines have the ability to complete any task in a given time interval for completion of task required intelligence, which is referred to as the AI-effect (McCorduck, 2004). According to Tesler’shypothesis states, “whatever has not been completed yet, with the help of AI is possible (Maloof, 2006). For example, recognition of optical character is possible by using an AI system (Levin & Pieraccini, 1992). With the advancement in AI technology, machines are able to understand human speech (Underwood, 1977) that competes for the utmost level in the strategic system including different computer games including chess, etc., operational cars and militant simulations.

Advancement in the research of AI into various sub-areas that frequently nosedive to connect with each other in a specific domain (Linn & Clancy, 1992). These various sub-fields are established on scientific concerns including specific objectives like modern robotics (Struijk, 2012), with the help of specific tools i.e is as logic, Artificial Neural Networks (ANN), deep theoretical changes. These various sub-areas have also been established on social aspects, mainly research institutions or particular researchers state of the artwork. However, various new different challenges associated with AI research with the passage of time such as reasoning, representation of information, preparation, planning, learning ability, language processing, observation and capability to relocate and control the objects (Kellman & Spelke, 1983). In this context, General Intelligence (GI) have the potential ability to resolve such issue and suitable alternative for the field's long-term goals (Voss, 2007).

Numerous methodologies mainly statistical approach and computational intelligence technique were extensively applied in the AI field. Various different tools are utilized in their field, including different types of exploration and numerical optimization, ANNs, and approaches established on statistical values, probability, and economics fields. The AI field attracts by computer science, information technology, fields of different branches of math, psychology, semantics, philosophy, and various different area of science. AI area was initiated on the hypothesis on human intelligence can accurate that suggested that machines might be finished to simulate it (Moravec, 1988). The advancement leads to logical opinions about the nature of the human mind and beliefs of making artificial things that offer human identical intelligence to the real world. Such associated problems have been identified by using fiction, myth, and philosophy (Morgan, 2000). Some of the researchers also suggested that AI might be dangerous to the human community, thereby need to be restricted progress or development (Woolgar, 1985). On the other hand, some researchers also believe that AI-based technologies might be producing mass-unemployment like previous technologies (Boyd & Holton, 2018). Moreover, AI technologies have experienced are surrection subsequent innovations in computer power, a huge amount of data processing, and theoretical interpretation. AI technologies have developed an indispensable portion of the industrial technologies that unravel various challenges in the area of computer science and operations research. The development of newer AI-based technologies has been one of the acute approaches of various sectors throughout the world. Several studies have been performed, however, the results of these approaches still need to be organized. In this context, ML with AI technologies is an essential tool for social development in various aspects.

Complete Chapter List

Search this Book:
Reset