Proceedings of Emerging Trends and Technologies on Intelligent Systems: ETTIS 2022 (Advances in Intelligent Systems and Computing, 1414)

This book presents best selected papers presented at the 2nd International Conference on Emerging Trends and Technologies on Intelligent Systems (ETTIS 2022) to be held from 22 – 23 March 2022 in online mode at C-DAC, Noida, India. The book includes current research works in the areas of artificial intelligence, big data, cyber-physical systems, and security in industrial/real-world settings. The book illustrates on-going research results, projects, surveying works, and industrial experiences that describe significant advances in all of the related areas. Preface Contents About the Editors EmotiSync: Music Recommendation System Using Facial Expressions 1 Introduction 2 Literature Survey 3 Existing System 4 Methodology 4.1 Face Detection 4.2 Emotion Recognition 4.3 Music Recommendation 5 Experimental Analysis 6 Result Analysis 7 Conclusion and Future Scope References Retrospective Review on Object Detection Approaches Using Boundary Information 1 Introduction 2 Object Detection 2.1 Boundary Detection 2.2 Edge Detection 2.3 Line Approximation Algorithm 3 Applications 4 Literature Review 5 Conclusion 6 Future Scope References Question Classification Based on Cognitive Skills of Bloom's Taxonomy Using TFPOS-IDF and GloVe 1 Introduction 2 Literature Review 3 Methodology 3.1 Question Dataset 3.2 Natural Language Processing 3.3 Feature Extraction 3.4 Classification 3.5 Evaluation 4 Implementation 4.1 TF-IDF Approach 4.2 TF-IDF + GloVe Approach 4.3 TFPOS-IDF + GloVe Approach 5 Results 5.1 Results of K-Nearest Neighbor (KNN, where K=3) 5.2 Results of Logistic Regression 5.3 Results of Random Forest Classifier 5.4 Results of AdaBoost Classifier 5.5 Results of Gradient Boosting Classifier 5.6 Results of XGBoost Classifier 5.7 Results of Support Vector Machine (SVM) 5.8 Comparison of Different Approaches Using the SVM Model 5.9 Overall Results 6 Future Work and Conclusion References Sign2Sign: A Novel Approach Towards Real-Time ASL to ISL Translation 1 Introduction 2 Related Work 3 Methodology 3.1 Labeling and Collection of the Dataset 3.2 Creating the Hand Histogram 3.3 Preprocessing the Input Hand Postures 3.4 Passing Pre-processed Images to the CNN 3.5 Predicting ASL Signs 3.6 Stringing Together Corresponding ISL Signs to Form Meaningful Videos 4 Results 5 Conclusions 6 Future Directions References Analysis of Patient Tuberculosis Tenet Death Reason and Prediction in Bangladesh Using Machine Learning 1 Introduction 2 Related Work 3 Dataset 3.1 Dataset Collection 3.2 Dataset Elaboration 4 Methods and Process 4.1 Dataset Scrubbing 4.2 Feature Importance Techniques Using Machine Learning Algorithms 4.3 Visualization of Algorithm Feature Selection 4.4 Visualization and Analysis of Regression Algorithm 4.5 Visualization and Analysis of Classification Algorithm 4.6 Prediction Accuracy of Algorithms 5 Discussion 6 Conclusion References Portable Electronic Tongue for Characterisation of Tea Taste 1 Introduction 2 Description of Membrane Sensor Electrode and Array 3 Description of the Embedded Electronics 4 Experimentation 4.1 Stability and Repeatability Check of the Membrane Sensors 4.2 Experimentation with Tea Samples 5 Analysis of Acquired Data 5.1 Principal Component Analysis (PCA) 5.2 Multiple Discriminant Analysis (MDA) 5.3 Probabilistic Neural Network (PNN) 5.4 Back Propagation—Multi Layer Perceptron (BP-MLP) 6 Results and Discussion 7 Conclusion References e-Visit Using Dynamic QR Code with Application Deep Linking Capability: Mobile-App-Based Solution for Reducing Patient's Waiting Time 1 Introduction 2 Related Work 3 Approach and Workflow 4 System Architecture 4.1 Features of QR-Based E-Visit 4.2 Department-wise QR Generation Process 4.3 Mobile App Visit Process 4.4 Hospital KPIs and Statistics 5 Conclusion and Future Work References Gunshot Detection and Classification Using a Convolution-GRU Based Approach 1 Introduction 2 Related Work 3 Methodology 3.1 Converting Audio to MFCCs 3.2 Constructing the Audio Dataset 3.3 Training a Deep Learning Model 4 Model 5 Results 6 Conclusion References Different Skin Tone Segmentation from an Image Using KNN for Sign Language Recognition 1 Introduction 2 Related Work 2.1 Color Space 3 Experiments 3.1 Dataset Description 3.2 Supervised Learning Algorithms Comparative Study 4 Conclusion References MuteMe—An Automatic Audio Playback Controller During Emergencies 1 Introduction 2 Related Works 3 Proposed Architecture 4 Implementation 4.1 Platform 4.2 Software Development 4.3 Machine Learning 5 Comparative Analysis 6 Conclusion References Chi-Square Top-K Based Incremental Feature Selection Model for BigData Analytics 1 Introduction 1.1 Motivation 1.2 Contributions 1.3 Organization 2 Related Works 3 Proposed System 3.1 Problem Statement 3.2 Objectives 4 Experimental Results and Discussion 5 Conclusions References M-Vahitaram: AI-Based Android Application for Automated Crowd Control Management in Bus Transport Service 1 Introduction 2 Literature Survey 3 Proposed System 4 Methodology 4.1 Database 4.2 M-Vahitaram Model 4.3 Implementation 4.4 Components 5 Results and Analysis 5.1 Analysis 6 Conclusion References Automatic Enhancement of Deep Neural Networks for Diagnosis of COVID-19 Cases with X-ray Images Using MLOps 1 Introduction 2 Technologies Used 3 Proposed Work 4 Method 4.1 Problem Formulation 4.2 Training the Model 5 Implementation of the Proposed Pipeline 5.1 Setting up the Docker Containers 5.2 Building the Jenkins Pipeline 6 Overall Performance Evaluation 7 Comparative Analysis with Related Works 8 Possible Potential Threats 9 Conclusion and Future Work References Big Data Disease Prediction System Using Vanilla LSTM: A Deep Learning Breakthrough 1 Introduction 2 Dataset 3 Literature Survey 4 Proposed Vanilla LSTM 5 Result and Discussion 6 Conclusion References Non-destructive Quality Evaluation of Litchi Fruit Using e-Nose System 1 Introduction 2 Materials and Methods 2.1 Sample Collection & Preparation 2.2 Sensor Array Optimization: (Feature Optimization) 2.3 Data Acquisition and Analysis 3 Experimental Results and Discussions 3.1 Sensor Array Optimization 3.2 Data Analysis 4 Conclusion References A Survey of Learning Methods in Deep Neural Networks (DDN) 1 Introduction 2 Challenges Before Deep Learning 3 Deep Learning Architectures 4 Types of Neural Networks 4.1 Deep Neural Network (DNN) 4.2 Recurrent Neural Network (RNN) 4.3 Challenges and Merits of DNN 4.4 Traditional ML Versus DL 5 Types of Learning Methods in Deep Learning 5.1 Supervised Learning 5.2 Semi-Supervised Learning 5.3 Unsupervised Learning 5.4 Reinforcement Learning (RL) 6 Comparison of Different Deep Learning Algorithms 7 Discussion and Conclusion References The Implementation of Object Detection Using Deep Learning for Mobility Impaired People 1 Introduction 2 Literature Review 3 Methodology 3.1 Implementing Object Detection Using the MobileNet SSD V3 on Open CV 3.2 Object Detection Using the Raspberry Pi V3 b + and the Pi Camera 4 Results 5 Conclusion 6 Future Scope References A Study on Deep Learning Frameworks for Opinion Summarization 1 Introduction 2 Deep Learning-Based Opinion Summarization Methods 2.1 Recurrent Neural Networks-Based Models 2.2 Convolutional Neural Networks-Based Models 2.3 Transformer-Based Models 2.4 Hybrid Models 2.5 Summary of Various Approaches 3 Evaluation Metrics 3.1 Rouge 3.2 Rouge-N 3.3 Rouge-L 3.4 Rouge-S 4 Datasets 4.1 Oposum 4.2 Space 4.3 Rotten Tomatoes 4.4 Opinosis 4.5 AmaSum 4.6 Yelp 5 Discussion and Future Works 6 Conclusion References Improvisation of Information System Security Posture Through Continuous Vulnerability Assessment 1 Introduction 2 Literature Review 3 The Continuous Vulnerability Assessment Security Dashboard 3.1 Asset and Network VA/PT 3.2 Mobile and Web Application VA/PT 3.3 ISMS Audit Module 3.4 Reporting Tool 3.5 Risk Score Index 4 Statistical Model of RSI in CVA Security Dashboard 5 Implementation and Evaluation 5.1 Comparative Analysis 6 Conclusions and Future Scope References Design and Development of Micro-grid Networks for Demand Management System Using Fuzzy Logic 1 Introduction 1.1 Scopes 1.2 Structure of the Paper 2 Background and Current Situation 2.1 Micro-networks 3 Demand Management 3.1 Electricity Demand and Micro-Grids 3.2 Prediction of Electricity Demand 3.3 Short-Term Demand Forecast 3.4 Method Regression 3.5 Series of Time 3.6 Approximation by Similarity Between days 3.7 Neural Networks 3.8 Expert System 3.9 Diffuse Logic 4 Proposed Work 4.1 Demand Prediction Based on Fuzzy Modeling 4.2 Fuzzy Takagi and Sugeno Models 4.3 Stability Analysis 4.4 Performance Analysis 4.5 Relevant Variables 5 Conclusion References Brain Tumor Detection Using Improved Otsu’s Thresholding Method and Supervised Learning Techniques at Early Stage 1 Introduction 2 Related Work 3 Methodology 3.1 Image Enhancement 3.2 Image Thresholding 3.3 Machine Learning Technique Applied 4 Experimental Design 4.1 Data Set Collection and Analysis 4.2 Performance Parameters 5 Results 6 Conclusion References Hyperspectral Image Prediction Using Logistic Regression Model 1 Introduction 2 Logistic Regression Model 2.1 Steps to Classify the Images Using Logistic Regression (LR) Model 2.2 Key Points of LR Model 3 Dataset and Features 4 Result and Discussion 5 Conclusion References Extractive Long-Form Question Answering for Annual Reports Using BERT 1 Introduction 2 Related Work 2.1 BERT 2.2 FinBERT 2.3 ELI5 Long Form Question Answering 2.4 BERTSerini 3 Dataset 4 Method 4.1 Extraction 4.2 TF-IDF 4.3 FinBERT Embeddings 4.4 Pre-trained and Fine-Tuned BERT 5 Results and Evaluation 6 Source Code 7 Conclusion References Endpoint Network Behavior Analysis and Anomaly Detection Using Unsupervised Machine Learning 1 Introduction 2 Data Collection and Generation of Dataset 2.1 Data Collection 2.2 Experiment Setup 3 Machine Learning Methods and Comparative Analysis 3.1 Data Preprocessing 3.2 Principal Component Analysis (PCA) 3.3 Machine Learning Methods for Anomaly Detection 4 Experiment and Results 5 Conclusion 6 Further Research References Handling Cold-Start Problem in Restaurant Recommender System Using Ontology 1 Introduction 2 Literature Review 3 Methodology 3.1 Creating an Ontology 3.2 Loading the Ontology in a Python Environment 3.3 Making the Program to Get User Preferences 4 Results and Evaluation 4.1 Prediction Rating Accuracy 4.2 Classification Accuracy 5 Conclusion and Future Work References An SVM-Based Approach for the Quality Estimation of Udupi Jasmine 1 Introduction 2 Literature Review 3 Methodology 4 Results and Discussion 5 Conclusion References Routing-Based Restricted Boltzmann Machine Learning and Clustering Algorithm in Wireless Sensor Network 1 Introduction 1.1 Wireless Sensor Networks 1.2 Restricted Boltzmann Machine Learning (RBM) Algorithm 2 Related Works 3 Proposed Methodology 3.1 Cluster Formation 3.2 RBMCA Process 3.3 Routing 4 Performance Analysis 5 Conclusion and Future Work References A Systematic Review on Underwater Image Enhancement and Object Detection Methods 1 Introduction 1.1 Restoration of Under-Water Images 1.2 Enhancement of Under-Water Images 2 Literature Review 2.1 Literature Survey on Underwater Image Enhancement 2.2 Literature Survey on Underwater Image Detection 3 Underwater Image Quality Evaluation 3.1 Subjective/Qualitative IQA 3.2 Objective/Quantitative IQA 4 Underwater Dataset 5 Applications 6 Discussion 7 Conclusion References IoT-based Precision Agriculture: A Review 1 Introduction 2 Data Acquisition Techniques 2.1 WSNs 2.2 Image Acquisition 3 Data Analysis 4 Technical Discussion 5 Conclusion References Enhancing the Security of JSON Web Token Using Signal Protocol and Ratchet System 1 Introduction 1.1 Problem Statement and Our Contributions 2 Related Works 3 Secure JSON Tokens 3.1 Double Ratchet Algorithm 3.2 Signal Protocol 4 Methodologies 5 Implementation 5.1 Dependencies 5.2 Initialization 5.3 Introduction Code 5.4 Ratchets 5.5 Sending and Receiving Messages 6 Conclusion and Future Works References Price Prediction of Ethereum Using Time Series and Deep Learning Techniques 1 Introduction 2 Related Works 3 Proposed Model 4 Methodology 4.1 Data Collection 4.2 Data Processing 4.3 Feature Selection 4.4 Train and Test Set 5 Experimental Results 5.1 Results 5.2 Loss Curve 5.3 Discussions 6 Conclusion References Light Weight Approach for Agnostic Optimal Route Selection 1 Introduction 1.1 Motivation 2 Related Work 2.1 Uninformed Search 2.2 Informed Search 3 System Model 4 Proposed Scheme 4.1 Reforming Data 4.2 One Leg Recommendation 4.3 Two-Leg Recommendation 4.4 Three Leg Recommendation 4.5 Four-Leg Recommendation 5 Performance and Comparison Analysis 6 Summary and Future Work References Index