Machine Intelligence : Computer Vision and Natural Language Processing

Machines are being systematically empowered to be interactive and intelligent in their operations, offerings. and outputs. There are pioneering Artificial Intelligence (AI) technologies and tools. Machine and Deep Learning (ML/DL) algorithms, along with their enabling frameworks, libraries, and specialized accelerators, find particularly useful applications in computer and machine vision, human machine interfaces (HMIs), and intelligent machines. Machines that can see and perceive can bring forth deeper and decisive acceleration, automation, and augmentation capabilities to businesses as well as people in their everyday assignments. Machine vision is becoming a reality because of advancements in the computer vision and device instrumentation spaces. Machines are increasingly software-defined. That is, vision-enabling software and hardware modules are being embedded in new-generation machines to be self-, surroundings, and situation-aware. Machine Intelligence emphasizes computer vision and natural language processing as drivers of advances in machine intelligence. The book examines these technologies from the algorithmic level to the applications level. It also examines the integrative technologies enabling intelligent applications in business and industry. Features Motion images object detection over voice using deep learning algorithms Ubiquitous computing and augmented reality in HCI Learning and reasoning in Artificial Intelligence Economic sustainability, mindfulness, and diversity in the age of artificial intelligence and machine learning Streaming analytics for healthcare and retail domains Covering established and emerging technologies in machine vision, the book focuses on recent and novel applications and discusses state-of-the-art technologies and tools. Cover Half Title Title Page Copyright Page Table of Contents List of Figures List of Tables Preface Contributors Abbreviations Chapter 1 A New Frontier in Machine Intelligence: Creativity 1.1 Introduction 1.2 A Short History of Computer Creativity 1.3 Artificial Intelligence and Creativity 1.4 Artificial Creativity: The Debate 1.5 Artificial Creativity and Copyright 1.6 Conclusion 1.7 Postscript References Chapter 2 Overview of Human-Computer Interaction 2.1 Introduction 2.2 The Aim of HCI 2.3 Factors in HCI 2.4 HCI Design Issues 2.4.1 Human-Machine Symbiosis 2.4.2 Human-Environment Interactions 2.4.3 Ethics, Privacy, and Security 2.4.4 Well-Being, Health, and Eudaimonia 2.4.5 Accessibility and Universal Access 2.4.6 Learning and Creativity 2.4.7 Social Organization and Democracy 2.5 HCI Implementation Issues 2.6 Important Aspects of HCI 2.7 Components of HCI 2.8 The Characteristics of HCI 2.9 HCI Principles and Best Practices 2.10 Design for HCI 2.10.1 HCI Design Approaches 2.10.2 Interface Testing Techniques 2.11 HCI Devices 2.12 HCI Tools and Technologies 2.13 HCI’s Eye-Tracking Technology 2.13.1 Eye-Tracking With Head Stabilization 2.13.2 Remote Eye-Tracking 2.13.3 Mobile Eye-Tracking 2.13.4 Embedded Or Integrated Systems 2.14 HCI’s Speech Recognition Technology 2.15 The Internet of Things (IoT) Technology 2.16 HCI’s Cloud Computing Technology 2.17 HCI Applications 2.18 Advantages of HCI 2.19 Disadvantages of HCI 2.20 Conclusion References Chapter 3 Edge/Fog Computing: An Overview and Insight Into Research Directions 3.1 Introduction 3.2 Edge/Fog Computing Basics 3.2.1 Edge Computing 3.2.2 Fog Computing 3.2.3 Understanding the Differences Between Edge and Fog Computing 3.2.4 Edge/Fog Computing Characteristics 3.2.4.1 Low Latency 3.2.4.2 Less Bandwidth Consumption 3.2.4.3 Mobility Support 3.2.4.4 Heterogeneity 3.2.4.5 Location Awareness 3.2.5 Benefits of Edge/Fog Computing 3.2.5.1 Cost 3.2.5.2 Speed 3.2.5.3 Scalability 3.2.5.4 Performance 3.2.5.5 Security 3.2.5.6 Reliability 3.2.5.7 Agility 3.3 Architecture of the Edge/Fog Computing 3.3.1 Basic Components of the Three-Tier Architecture 3.3.1.1 End Device Layer/Physical Layer 3.3.1.2 Fog/Edge Layer 3.3.1.3 Cloud Layer 3.4 Applications of Edge/Fog Computing 3.4.1 Healthcare Industry 3.4.2 IoT Applications 3.4.3 Augmented Reality (AR) 3.4.4 Banking and Finance Industry 3.4.5 Manufacturing Industries 3.4.6 Automobile Industry 3.4.7 Summary 3.5 Challenges and Research Directions 3.5.1 Increased Complexity 3.5.2 Privacy and Security 3.5.3 Task Scheduling/Offloading 3.5.4 Power Consumption 3.5.5 Data Management 3.5.6 Quality of Service 3.5.7 Multi-Characteristics Fog Design System 3.6 Conclusion References Chapter 4 Reduce Overfitting and Improve Deep Learning Models’ Performance in Medical Image Classification 4.1 Introduction 4.2 Deep Convolutional Neural Networks (DCNNs) 4.3 Medical Image Classification Using Transfer Learning 4.3.1 Challenges When Using Transfer Learning in Medical Image Classifications 4.3.2 Applications of Transfer Learning in Medical Image Classifications 4.4 Overfitting Prevention Techniques 4.4.1 Weight Regularization 4.4.2 Activity Regularization 4.4.3 Adding Dropout Layers 4.4.4 Noise Regularization 4.4.5 Stop Training With the Early Stopping Method 4.5 Conclusion References Chapter 5 Motion Images Object Detection Over Voice Using Deep Learning Algorithms 5.1 Introduction 5.1.1 Self-Driving Vehicles 5.1.2 Levels of Autonomy 5.1.3 Digital Image Processing 5.1.3.1 Image Pre-Processing 5.1.4 Deep Learning 5.1.5 Convolution Neural Networks 5.1.6 Computer Vision 5.1.6.1 Applications of Computer Vision 5.1.7 Object Detection 5.1.8 YOLO: You Only Look Once 5.2 Literature Review 5.3 Proposed Methodology 5.3.1 Proposed Architecture 5.3.2 Labelled Input Image 5.3.3 Stages of the Proposed Method 5.3.4 YOLOv4 Architecture 5.3.4.1 Intersection Over Union (IoU) 5.3.4.2 MAP 5.3.4.3 Precision 5.3.4.4 Recall 5.3.4.5 Non-Maximum Suppression (NMS) 5.3.5 Google Text-To-Speech 5.3.6 YOLOv5 Models’ Approach to Checking Performance 5.3.6.1 Design of the Proposed Method 5.4 Experimental Results 5.4.1 Detection of MAP Performance 5.4.2 Experimental Outcome Results 5.4.3 Predictions 5.5 Conclusion References Chapter 6 Diabetic Retinopathy Detection Using Various Machine Learning Algorithms 6.1 Introduction 6.2 Background 6.2.1 Proliferative Diabetic Retinopathy 6.3 Applicability 6.4 Diabetic Retinopathy Detection 6.4.1 Vertical and Horizontal Flip in Image Augmentation 6.4.2 Random Rotation Augmentation 6.4.2.1 Transition Layer 6.4.3 DenseNet for Semantic Segmentation 6.4.4 Dataset Description 6.5 Conclusion Note References Chapter 7 IIoT Applications and Services 7.1 Introduction 7.1.1 The Need for IoT 7.1.2 Integration of IoT and Technologies 7.2 Components of IoT 7.3 Programming Software for IoT 7.3.1 Python for IoT 7.3.2 Python for Backend Development 7.3.3 IDEs for Internet of Things Development (IoT) 7.4 IoT Hardware 7.5 Industrial IoT (IIoT) 7.5.1 Applications of IIoT 7.5.2 IIoT Sensors 7.5.3 IoT Sensors for Industrial Automation Solutions 7.6 Artificial Intelligence and IoT 7.7 IIoT Start-Ups in India 7.8 Challenges in Securing IoT in India 7.9 Policies and Regulations for Promoting IoT in India 7.9.1 Recommendations for IoT Devices 7.10 Applications of IoT 7.11 Use-Cases in Industrial IoT 7.11.1 Predictive Maintenance 7.11.2 Smart Metering 7.11.3 Location Tracking 7.11.4 Location Services 7.11.5 Remote Quality Monitoring 7.11.6 Supply Chain Management and Optimization 7.12 Future of IIoT 7.13 Conclusion References Chapter 8 Design of Machine Learning Model for Health Care Index During COVID-19 8.1 Introduction 8.2 Literature Review 8.3 Time Series Data 8.4 Development of the ARIMA Model 8.5 Conclusion References Chapter 9 Ubiquitous Computing and Augmented Reality in HCI 9.1 Introduction 9.2 Ubiquitous Computing (UC) 9.2.1 UC’s History 9.2.2 Characteristics of UC 9.2.3 UC’s Layers 9.2.4 UC Types 9.3 UC Devices 9.3.1 Smartwatches 9.3.2 Smart Rings 9.3.3 Advanced Medical Wearables 9.3.4 Smart Earphones 9.3.5 Smart Clothing 9.4 UC’s Applications 9.4.1 Healthcare Industry 9.4.2 Accessibility 9.4.3 Learning 9.4.4 Logistics 9.4.5 Commerce 9.4.6 Games 9.5 Advantages of UC 9.6 Disadvantages of UC 9.7 Augmented Reality (AR) 9.7.1 AR’s History in a Nutshell 9.7.2 Characteristics of AR 9.7.3 AR Types 9.7.3.1 Marker-Based AR 9.7.3.2 Markerless AR 9.7.3.3 Location-Based AR 9.7.3.4 Superimposition AR 9.7.3.5 Projection-Based AR 9.8 AR Devices 9.8.1 Microsoft HoloLens 9.8.2 MagicLeap One 9.8.3 Epson Moverio 9.8.4 Google Glass Enterprise Edition 9.9 AR Applications 9.9.1 In the Military 9.9.2 3D Animals 9.9.3 Fashion 9.9.4 Gaming 9.9.5 Coloring Books 9.9.6 Obstetrics 9.9.7 Architecture 9.9.8 Sports 9.10 Advantages of AR 9.11 Disadvantages of AR 9.12 Conclusion Acknowledgments References Chapter 10 A Machine Learning-Based Driving Assistance System for Lane and Drowsiness Monitoring 10.1 Introduction 10.2 Literature Review 10.2.1 System Review 10.2.2 Lane Detection Techniques 10.2.3 Robust Lane Detection in Low Light 10.2.4 Requirement 10.2.5 Use Case Diagram 10.2.6 Process Flow 10.3 Framework for Performance Analysis 10.3.1 Working On Drowsiness Detections 10.4 Image Capturing 10.4.1 Edge Detection 10.4.2 Feature Extraction Frontal Face 10.4.3 Grayscale Conversion 10.4.4 Score Calculation 10.5 Proposed Model for Lane Detection System 10.5.1 Comparing the Accuracy of the Lane Detection System 10.5.2 Implementation 10.5.2.1 System Design Approach 10.6 Computational Results 10.6.1 Accuracy Achieved 10.6.2 Assumptions 10.7 Conclusion References Chapter 11 Prediction of Gastric Cancer From Gene Expression Dataset Using Supervised Machine Learning Models 11.1 Introduction 11.2 Methodology 11.2.1 Dataset Description 11.2.2 Classification Techniques 11.2.2.1 Logistic Regression (LR) 11.2.2.2 Decision Tree (DT) 11.2.2.3 Naïve Bayes (NB) 11.2.2.4 K-Nearest Neighbor (KNN) 11.3 Results and Discussion 11.4 Conclusion References Chapter 12 Sewer Pipe Defect Detection in CCTV Images Using Deep Learning Techniques 12.1 Introduction 12.2 Related Work 12.3 Proposed Methodologies 12.3.1 Faster R-CNN for Object Detection 12.3.1.1 Convolutional Layer 12.3.1.2 Feature Map 12.3.1.3 Region Proposal Network (RPN) 12.3.1.4 Region of Interest (ROI) Pooling 12.3.1.5 Max Pooling Layer 12.3.2 System Architecture 12.3.2.1 Labeled Input Image 12.3.2.2 Identify Defects Using Faster R-CNN 12.3.2.3 Training With ZF/VGG/RESNET50 Networks 12.3.2.4 RGB Image Converted to HSV Image 12.3.3 Crack Detection 12.4 Experimental Results 12.4.1 Performance Analysis 12.4.2 Measure of Effectiveness 12.4.3 Experimental Outcome Results 12.5 Conclusion References Chapter 13 Learning and Reasoning Using Artificial Intelligence 13.1 Introduction 13.1.1 Human Intelligence Based On the Psychological View 13.1.2 Types of Artificial Intelligence Based On Functionality 13.1.2.1 Type 1 Category 13.1.2.2 Type 2 Category 13.1.3 Types of AI Based On Technology 13.1.4 Types of Intelligence 13.1.5 Structure of Intelligent Agents 13.1.6 How Intelligent Agents Work 13.1.7 Intelligent Agent Applications 13.1.8 Artificial Intelligence in Everyday Application 13.2 Growing Use of AI in Online Applications 13.2.1 E-Commerce 13.2.2 E-Learning Tools 13.2.3 Conducting Auctions 13.2.4 Travel Reservations 13.2.5 Online Stock Trading 13.2.6 Electronic Banking 13.2.7 Advertising and Marketing 13.2.8 Customer Service 13.3 AI and Security and Surveillance 13.4 AI and Medical Image Processing 13.4.1 Cardiovascular Abnormalities 13.4.2 Musculoskeletal Injury 13.4.3 Neurological Diseases 13.4.4 Thoracic Condition and Complications 13.5 Advantages of AI 13.5.1 Automation 13.5.2 Smart Decisions By AI 13.5.3 Increased Customer Experience 13.5.4 Medical Support System 13.5.5 Data Analysis 13.5.6 Stability of Business 13.5.7 Managing Repetitive Tasks 13.5.8 Minimizing Errors 13.6 Training AI 13.6.1 Success of AI Training 13.6.2 Train, Test and Maintain AI and Machine Learning Models 13.7 Conclusion References Chapter 14 A Novel Auto Encoder-Network-Based Ensemble Technique for Sentiment Analysis Using Tweets On COVID-19 Data 14.1 Introduction 14.2 Background and Related Work 14.2.1 Classification of Sentiments 14.2.2 Subjectivity Classification 14.2.3 Opinion Spam Identification 14.2.4 Detection of Language Implicitly 14.2.5 Extraction of Aspects 14.2.6 Datasets Associated With Sentiment Analysis 14.2.7 Approaches to Sentiment Analysis 14.2.8 Levels of Sentiment Analysis 14.3 Research Methodology 14.3.1 Data Extraction 14.3.2 Data Pre-Processing 14.3.3 Polarity Classification 14.3.4 Tweet Classification 14.3.5 Auto-Encoder 14.3.6 Adam Optimization 14.3.7 Ensemble Techniques for Sentiment Analysis 14.3.8 Decision Tree Classifier 14.3.9 Gradient Boosting Classifier 14.3.10 Logistic Regression 14.3.11 Genetic Algorithm 14.3.12 Support Vector Machine 14.3.13 Dataset Visualization 14.3.14 Word Cloud 14.4 Packages/Libraries of Python 14.4.1 Mlrose: Machine Learning, Randomized Optimization and Search 14.4.2 TfidfVectorizer 14.4.3 Train_test_split 14.4.4 Logistic Regression 14.5 Conclusion Acknowledgments References Chapter 15 Economic Sustainability, Mindfulness, and Diversity in the Age of Artificial Intelligence and Machine Learning 15.1 Use of AI and ML in Traditional Industry 15.2 Use of AI to Create an Agricultural Database 15.3 Use of AI to Be Sensitive to Users’ Emotions 15.4 Livelihood Training and a Caution for CSR Funders 15.5 Organic Farming and Sustainable Livelihoods 15.5.1 Agarbatti Smoke and Tobacco Smoke 15.6 The Role of Banks and Financial Institutions 15.7 Diversity 15.8 Mindfulness 15.9 Conclusion References Chapter 16 Adopting Streaming Analytics for Healthcare and Retail Domains 16.1 Introduction 16.2 Healthcare Data Sources and Basic Analytics 16.2.1 Patient Data 16.2.2 Medical Imaging Records 16.2.3 Sensing Device Data 16.2.4 Mining Clinical Notes 16.2.5 Community Data Analysis 16.3 Retail Domain 16.3.1 Retail Industry 16.3.2 Customer Satisfaction in Real-Time Analysis 16.3.3 Technological Applications in the Retail Domain 16.3.3.1 Inventory Tracking 16.3.3.2 Customer Service 16.3.3.3 Data Warehousing 16.3.4 Retail Data Sources and Analysis 16.4 Different Ways of Streaming Analytics 16.5 Real-Time Streaming Analytics in Healthcare Use Cases 16.5.1 Data Stream Computing in Healthcare 16.5.1.1 Stream Computing Technology: Apache 16.5.1.2 Healthcare Analytics With Big Data: A Generic Architecture 16.6 Medical Signal Analysis 16.7 Big Data Analytical Signal Processing 16.8 Big Data Analytics in Genomics 16.8.1 Securing Genomic Data 16.8.2 Privacy 16.8.3 Data Ownership 16.9 Conclusion References Index Machines are being systematically empowered to be interactive and intelligent in their operations, offerings. and outputs. There are pioneering Artificial Intelligence (AI) technologies and tools. Machine and Deep Learning (ML/DL) algorithms, along with their enabling frameworks, libraries, and specialized accelerators, find particularly useful applications in computer and machine vision, human machine interfaces (HMIs), and intelligent machines. Machines that can see and perceive can bring forth deeper and decisive acceleration, automation, and augmentation capabilities to businesses as well as people in their everyday assignments. Machine vision is becoming a reality because of advancements in the computer vision and device instrumentation spaces. Machines are increasingly software-defined. That is, vision-enabling software and hardware modules are being embedded in new-generation machines to be self-, surroundings, and situation-aware.Machine Intelligence: Computer Vision and Natural Language Processing emphasizes computer vision and natural language processing as drivers of advances in machine intelligence. The book examines these technologies from the algorithmic level to the applications level. It also examines the integrative technologies enabling intelligent applications in business and industry.Features: Motion images object detection over voice using deep learning algorithms Ubiquitous computing and augmented reality in HCI Learning and reasoning in Artificial Intelligence Economic sustainability, mindfulness, and diversity in the age of artificial intelligence and machine learning Streaming analytics for healthcare and retail domains Covering established and emerging technologies in machine vision, the book focuses on recent and novel applications and discusses state-of-the-art technologies and tools.