Cover Page
image

Smart and Sustainable Intelligent Systems

Homogeneous Random Motions and their Applications

Namita Gupta

Prasenjit Chatterjee

Tanupriya Choudhury

images

Dedication

Prof. Namita Gupta would like to dedicate this book to her husband Mr. Arvind Goyal and her sons Anish Goyal and Ansh Goyal for their patience and support. She expresses extreme gratitude to Dr N. K. Garg (Founder and Chief Advisor, MATES), Prof. Neelam Sharma (Director MAIT) and Prof. S. S. Deswal (Dean Academics, MAIT) for their valuable guidance and encouragement.

Dr. Prasenjit Chatterjee would like to dedicate this book to his late grandparents, father Late Dipak Kumar Chatterjee, his mother Mrs. Kalyani Chatterjee, beloved wife Amrita and his little angel Aheli.

Dr. Tanupriya Choudhury would like to dedicate this book to his late grandparents, his father Sri Mrigendra Choudhury, his mother Mrs. Minakshi Choudhury, beloved wife Rituparna and his son Rajrup and his respected guides, his research collaborators, friends, scholars, most beloved students and colleagues.

Preface

This book covers emerging computational and knowledge transfer approaches, optimizing solutions in varied disciplines of science, technology and healthcare. The idea behind compiling this work is to familiarize researchers, academicians, industry persons and students with various applications of intelligent techniques for producing sustainable, cost-effective and robust solutions of frequently encountered complex, real-world problems in engineering and science disciplines.

The chapters include the list of topics that spans all the areas of smart intelligent systems and computing such as: Data Mining with Soft Computing, Evolutionary Computing, Quantum Computing, Expert Systems, Next Generation Communication, Blockchain and Trust Management, Intelligent Biometrics, Multi-Valued Logical Systems, Cloud Computing and security etc. An extensive list of bibliographic references at the end of each chapter guides the reader to probe further into application area of interest to him/her.

Organization of the Book

The complete book is organized into 36 Chapters. A brief description of each of the chapters is presented as follows:

Chapter 1 proposes to improve the residual block architecture to define a modified residual dense block with the addition of the batch-normalization layer and secondly & improvised the Perceptual loss function according to our GAN. After these two changes, the authors introduce a new GAN based architecture for the Single Image Super-Resolution task for higher up-sampling levels.

Chapter 2 proposes a solution to google landmark recognition challenge 2019 hosted on Kaggle and an intuition for facilitating a tour guide recommender engine and visualization using method attempts to predict landmarks using CNN, pre-trained with VGG16 neural network used with transfer learning from ImageNet.

Chapter 3 suggests that Neural Network Models specifically CNN performed way better than other models with some fine-tuning for classification of 200 species of birds with an accuracy of 86.5%.

Chapter 4 proposes the use of Advance Image Processing Techniques and OpenCV functions to detect lanes on a public road along with calculating the radius of curvature of the lane and vehicle position in respect to the road lane i.e. central offset. In this paper, the front facing camera on the hood of the car is used for recording the video of the road in front of the car and feeding that video in the algorithm for better predictions of the road area. Used techniques like Search from Prior and Sliding window to create a more efficient and accurate algorithm better than previous approaches.

Chapter 5 proposes a methodology to detect human and animated faces in real time video and images. The facial expression would then be classified six basic emotions- happy, surprise, fear, anger, sad, disgust and a neutral face.

Chapter 6 proposes a system that does not segment the input of images, but rather the layers extract relevant features from the scanned images fed as input. Compared with previous systems for handwritten text recognition, the given architecture is end-to-end trainable and does not require different components to be trained separately. It naturally handles sequences in random lengths, involving no horizontal scale normalization or character segmentation. The model is smaller yet effective, thus, more practical for application in real-world scenarios.

Chapter 7 presents a machine learning modelled system able to analyse the image of an automobile captured by a camera, detect the registration plate and identify the registration number of the automobile. The algorithm also accepts live feed videos and pre-recorded videos, which are broken into frames. This system was made to solve security problems which exist in residential buildings, societies, parking areas and other institutions and areas where the pre-existing security systems cannot be installed.

Chapter 8 presents the finding of the best algorithm that can be used to predict the disease or chances that the disease can occur in the person.

Chapter 9 build a model which is combination of CNN model classification problem (to predict whether the subject has brain tumor or not) & Computer Vision problem (to automate the process of brain cropping from MRI scans). VGG-16 model architecture is used for feature extraction, which along with other features is fed as input to an Artificial Neural Network classifier through transfer learning. Binary Classifier classifies the input images as either 0 (No tumor) or 1 (Tumor exists).

Chapter 10 discusses the challenges faced in understanding the gestures of deaf and mute people in India and identified the need for a proper translator arises.

Chapter 11 designs a heterogeneous 3-Dimensional mesh (or network) to co-train 3 dimensional medical images dataset so as to make a series of pre-trained classifiers.

Chapter 12 explores a simpler approach to recommender systems that are fairly easy to use for small scale e- retail websites.

Chapter 13 discusses the recommendations to users on the basis of various user preferences while interacting with a social media platform, such as the gender, age, number of meetups, the skills and ratings of user profiles, image quality. All are taken into consideration to provide the ultimate user recommendation possible to make the software more relevant and user centric with recommendations strictly based on the way a given user interacts with the platform.

Chapter 14 discusses and compares the various machine learning algorithms applied to predict the literacy rate of various states.

Chapter 15 discusses an approach for video to video translation using various poses generated in the frames of video for translation. The approach makes use of Pose Generation Convolutional Neural Network to synthesize arbitrary poses from source videos and train the pix2pix - DCGAN which is a conditional generative adversarial network consisting of multi scale discriminator and generator for target video frames generation. It uses PatchGAN loss, VGG loss and Feature Matching Loss function for improving and optimizing models. The presented approach provides compelling results of the generated DCGAN model with the discriminator loss of 0.0003 and generator loss of 5.8206.

Chapter 16 compares different classification and boosting algorithms like Count Vectorizer with xG Gradient Boosting, TF-IDF Vectorizer with xG Gradient Boosting, Logistic Regression, and Random Forest.

Chapter 17 discusses development of traditional pixel-based methods and ends with the evolution of the latest object-based change detection techniques. LANDSAT and PALSAR images are used to represent the changes developed in the land use/land cover using pixel-based and object-based approaches.

Chapter 18 presents a study of four different machine learning algorithms namely J48, JRip, Random Forest and Naive Bayes, to detect three types of bad smells God Class, Long Method and Feature Envy. The results demonstrated that the machine learning algorithms achieved high accuracy with the validation method of 10-fold cross-validation.

Chapter 19 discusses the work done by the different researchers for identification of the negation’s cues and their scope. Chapter is organized according to the different feature selection methods employed and how different researchers contributed to this.

Chapter 20 describes the methodology and experimental setup used for the generation and development of bilingual speech corpus. Continuous and spontaneous (discrete) speech samples from both languages are collected on different mobile phones in a real-time environment, unlike studio environments. A brief comparative study of 18 readily available multilingual speech corpus developed for Indian languages is made against the proposed corpus. An annotation scheme is discussed to carry out further study how the recognition rate varies on the basis of language, device, text, and utterance.

Chapter 21 focuses on building an Intelligent Intrusion Detection System utilizing a blend of Nature Inspired Heuristics and Automated Machine Learning. The study applies Evolutionary Algorithms for feature selection as well as Hyperparameter Optimization. Moreover, the research explores Bayesian Search for Neural Architecture Search to estimate the ideal architecture of an artificial neural network (ANN).

Chapter 22 introduces the concept of distributed ownership of any digital asset (NFT) amongst many people in the form of percentage shares. Distributed NFT (dNFT), holds the properties of NFTs as well as it can be traded in the form of percentage shares as in real world. Each digital asset in this model is validated by the validators who act as a trust entity in the system. Hence, it is authentic, verifiable and acts as real market place for all types of digital assets.

Chapter 23 compares 5 most popular blockchain platforms on the basis of 21 different attributes concluded with a summary of different platforms and some suggestions for most commonly and widely used blockchain platforms.

Chapter 24 discusses how the installation of smart bins will contribute towards an enhanced waste management system that will create a circular economy coupled with evolving production and consumption behavior while minimizing the environmental impact.

Chapter 25 discusses the structure of IDS; different types of intrusion detection techniques and various types of attacks and compare various intrusion detection systems based on techniques used, various parameters of detection performance and their use in different domains.

Chapter 26 features increasing demand for green communication technology in telecom and IT industry for better energy efficiency.

Chapter 27 discusses the integration of IoT with cloud, edge and fog computing; challenges associated with it and its applications.

Chapter 28 discusses the importance of privacy in telecommunication systems, the cheap adversary actions taken to generate a social profile, targets a vulnerability in GSM mobile architecture using open source software and software-defined radio, and discusses the awareness of the existing vulnerabilities among the public. It also predicts the security awareness rating for colleges using ridge regression and binary classification.

In Chapter 29, a novel Consumer Oriented Trust Model in E-Commerce is discussed. Proposed model has divided into five levels with five components.

Chapter 30 discusses various data mining techniques that can be used for making profitable business management decisions.

Chapter 31 categorizes data deduplication techniques applied on stored data in distributed computing condition to provide security when database size expands. In this study the authors elaborated the coordinating procedures of scrambled information and open difficulties for information deduplication.

Chapter 32 proposes an approach towards procedural generation of music which can be used by a person without any musical knowledge. This can be used to create recreational music according to a person’s disposition. It can also be used in professional capacity to make music of different genres by content creators, game developers, animators and/or musicians to quickly create music loops, ambient/background sounds, gameplay music or even film scores with an interface providing control over each parameter to generate limitless variations.

Chapter 33 presents an approach that tackles photographs which have been morphed by using Photoshop alone using a neural network trained on real face images and their fake counterparts which have been manually generated by us using automation in Photoshop.

Chapter 34 presents review of various SQL injection approaches with the methods available to counter these attacks.

Chapter 35 discusses Futuristic Communication technology, Smart Energy Communication, Ubiquitous Communication and how Green technology sustainably affect environment and human life.

Chapter 36 includes survey on three types of cloud services: Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS) and Infrastructure-as-a service (IaaS).

The Editors are grateful to the many chapter contributors, acknowledged international authorities in their respective fields, whose contributions constitute the substance of this book. They have made time in their very busy lives to devote attention to the generation of their exceptional manuscripts. Appreciation is expressed to Mr Alok Sharma, Mr Anupam Kumar, Mr Ashish Sharma, Dr Pooja Gupta, Ms Ruchi Goel, Mr Yogesh Sharma and Dr Farzil Kidwai for helping in the manuscripts review process.

Acknowledgement

The editors express their sincere gratitude to all the chapter authors who contributed their time and expertise to this book.

The editors also wish to acknowledge the valuable contributions of all the reviewers who extended their supports for improvement of quality, coherence, and content presentation of chapters.

The editors express their gratitude and sincere thanks to Scrivener Publishing for providing such a great opportunity and extending cooperation and support in every facet.

Last but not the least, the editors are extremely thankful to all readers and look forward to receive constructive feedback and suggestions.

Part 1
MACHINE LEARNING AND ITS APPLICATION

Part 2
DEEP LEARNING AND ITS APPLICATION

Part 3
SECURITY AND BLOCKCHAIN

Part 4
COMMUNICATION AND NETWORKS

Part 5
LATEST TRENDS IN SUSTAINABLE COMPUTING TECHNIQUES