Advances in Speech and Music Technology: Proceedings of FRSM 2020 (Advances in Intelligent Systems and Computing)

Conference Organization
Keynote Talks
Preface
Contents
About the Editors
Vocal and Instrumental Music Processing
Music Signal Processing: A Literature Survey
	1 Introduction
		1.1 Indian Classical Music
		1.2 Carnatic Concert
		1.3 Classification of Indian Musical Instruments
	2 Areas of Research
		2.1 Source Separation
		2.2 Emotion Recognition
		2.3 Raga Classification
		2.4 Tala Classification
		2.5 Intonation and Rhythmic Analysis
		2.6 Tonic Identification
		2.7 Musical Note Representation
		2.8 Music Therapy
	3 Conclusion and Future Work
	References
Effects of Vocal Loading on Singing Power Ratio and Singer’s Formant with Indian Heavy Metal Vocalists
	1 Introduction
	2 Method
		2.1 Subjects
		2.2 Procedure
		2.3 Analysis
	3 Results and Discussion
	4 Conclusion
	References
Audio Quality Control Method Based on ASS (Audio Secret Sharing)
	1 Introduction
	2 Literature Review
	3 Proposed Method
	4 Analysis of Experiment Result
		4.1 Database Used for the Experiment
		4.2 Performance Measurements
		4.3 Experiment Result
		4.4 Significance of Neutral Range
		4.5 Comparison with Existing ASS Techniques
	5 Conclusion
	References
Noise Removal from Audio Using CNN and Denoiser
	1 Introduction
	2 Relevant Work
	3 Implementation
		3.1 Module 1. Pre-processing
		3.2 Module 2. Training
		3.3 Module 3. Noise Removal
	4 Results
	5 Conclusion
	References
Sine-Wave Speech as Preprocessing for Downstream Tasks
	1 Introduction
	2 Sine-Wave Representation of Speech
	3 Experiments and Results
		3.1 Dataset Used
		3.2 Experimental Studies
		3.3 Results
	4 Discussion
	5 Conclusions and Future Work
	References
Style of Vocal Singers in Indian Classical Music: Timbre Approach
	1 Introduction
	2 Results and Discussion
	3 Conclusion
	References
Style Identification of Vocal Singers in Indian Classical Music Using Meend and Andolan
	1 Introduction
	2 Meend
	3 Andolan
	4 Conclusion
	References
Computational Music and Speech
Speech Based Emotion Detection Using R
	1 Introduction
	2 Existing Algorithms for Emotion Detection
	3 Methodology
		3.1 DataSet
		3.2 Proposed Algorithm of the Emotion Detection Model
	4 Analysis and Results
	5 Conclusion
	References
Vocalist Identification in Audio Songs Using Convolutional Neural Network
	1 Introduction
	2 Related Work
		2.1 Machine Learning Approaches in Music Processing
		2.2 Baseline for Artist Classification
		2.3 Audio Representation as Spectrograms
		2.4 Convolutional Neural Networks
	3 Methodology
		3.1 Dataset
		3.2 Pre-processing of Audio Signals
		3.3 Architecture of the CNN Model
	4 Results and Discussion
	5 Conclusions and Future Work
	References
Swaragram: Shruti-Based Chromagram for Indian Classical Music
	1 Introduction
	2 Feature Extraction
		2.1 Tonic Identification
		2.2 Relative Swara Frequency
		2.3 Logarithmic Frequency Pooling Based on Shruti
		2.4 Swara Binning
	3 Toolbox
	4 Application for Indian Classical Music
		4.1 Raga Classification
	5 Conclusions and Future Work
	References
Machine Learning Approach for Audio Surveillance Using R
	1 Introduction
	2 Review of Existing Speech Recognition Methods
		2.1 Feature Extraction
		2.2 Classification
		2.3 DataSet
	3 Methodology
		3.1 Algorithm for Speaker Identification
		3.2 Algorithm for Gender Identification
	4 Analysis and Results
	5 Conclusion
	References
An Artificial Intelligence-Based Approach Towards Segregation  of Folk Songs
	1 Introduction
	2 Dataset
	3 Proposed Methodology
		3.1 Feature Extraction
		3.2 Ensemble Learning-Based Classification
	4 Result and Discussion
	5 Conclusion
	References
Shruti Detection Using Machine Learning and Sargam Identification for Instrumental Audio
	1 Introduction
	2 Preprocessing
	3 Shruti Detection
	4 Sargam Identification
		4.1 Frequency Identification
		4.2 Frequency Mapping to Swara
	5 Results
	6 Conclusions
	References
Addressing the Recitative Problem in Real-Time Opera Tracking
	1 Introduction
	2 Data Description
	3 Baseline Tracking Evaluation
	4 Specific Features for the Recitative Problem
		4.1 Feature Search Space
		4.2 Optimization on Fischer Recitatives
		4.3 Validating on Manacorda Recitatives
	5 Combining Trackers
		5.1 Combination Strategies
		5.2 Experiments
	6 Discussion and Conclusion
	References
Music Information Retrieval
IFSC: A Database for Indian Folk Songs Classification
	1 Introduction
	2 Related Work
	3 The IFSC5 Corpus
	4 Feature Representation
	5 Experiments
		5.1 Methodology
		5.2 Classifier
		5.3 Performance Evaluation
		5.4 Results and Discussion
	6 Conclusion and Future Directions
	References
North Indian Classical Music Tabla Tala (Rhythm) Prediction System Using Machine Learning
	1 Introduction
	2 State of the Art
	3 The Method
	4 Experiments and Results
	5 Conclusion and Future Scope
	Bibliography
Perception of Similarity and Dissimilarity in Hindustani Classical Music
	1 Introduction
	2 Thaat and Ragang Classification in Hindustani Classical Music
	3 Method
		3.1 Study 1
		3.2 Study 2
	4 Results and Discussion
		4.1 Study 1
		4.2 Study 2
	5 Conclusion and Future Work
	References
Analytical Comparison of Classification Models for Raga Identification in Carnatic Classical Audio
	1 Introduction
		1.1 A Glimpse into Indian Classical Music and Its Technical Aspects
	2 Related Work
	3 Methodology
		3.1 Preprocessing
		3.2 Feature Extraction
	4 Analysis of the Classification Models Used
		4.1 Artificial Neural Network (ANN)
		4.2 Long Short-Term Memory(LSTM)
		4.3 XGBoost
	5 Results
	6 Conclusion and Future Scope
	References
Multimodal Sentiment Analysis of Rabindra Sangeet Through Machine Learning Techniques
	1 Introduction
	2 Related Work
	3 Dataset
	4 Methodology
		4.1 Feature Extraction
		4.2 Pre-processing of Data
	5 Model and Experimental Results
		5.1 Lyrics Data
		5.2 Audio Data
	6 Performance of the Classification Models
	7 Conclusion and Future Scope of the Work
	References
Music Genre Classification with Convolutional Neural Networks and Comparison with F, Q, and Mel Spectrogram-Based Images
	1 Introduction
	2 Literature Review
	3 Proposed Methodology
		3.1 Dataset Collection
		3.2 Spectrogram Generation
		3.3 Convolutional Neural Networks (CNNs)
	4 Result Analysis
	5 Conclusion and Future Work
	References
A Simple Statistical Comparison of Bageshree and Bhimpalasi
	1 Introduction
	2 Methodology
	3 Experimental Results
	4 Discussion
	5 Conclusion
	References
Sound Analysis and Cognition
Effects of Musical Training on Auditory Spatial Processing Abilities: A Psychoacoustical and Perceptual Study
	1 Introduction
		1.1 Need for the Study
		1.2 Aim and Objectives
	2 Method
		2.1 Participants
		2.2 Informed Consent and Ethical Guidelines
		2.3 Informed Consent and Ethical Guidelines
		2.4 Informed Consent and Ethical Guidelines
	3 Results
	4 Discussion
		4.1 Outcomes of Musical Training on Temporal (ITD Thresholds) and Intensity Correlates (ILD Thresholds) of Spatial Hearing
		4.2 Outcomes of Musical Training on Composite Score (VASI Test) and Perceptual Scale (SSQ) of Spatial Hearing
		4.3 Comparison of the Pattern of Spatial Errors Between Musicians and Non-musicians
	5 Conclusions
	References
Effect of Transformation of Vowel Duration on Speech Intelligibility in Children with Hearing Loss
	1 Introduction
	2 Method
		2.1 Experiment 1
		2.2 Experiment 2
	3 Results and Discussion
		3.1 Experiment 1
		3.2 Experiment 2
	4 Conclusion
	References
The Influence of Music on Image Making: An Exploration of Intermediality Between Music Interpretation and Figurative Representation
	1 Introduction
	2 Research Questions
	3 Method
		3.1 Participants
		3.2 Stimuli
		3.3 Materials
		3.4 Procedure
		3.5 Analytical Strategy
	4 Results
	5 Discussion
	6 Conclusion
	References
Comparison of Neural Network Architectures for Speech Emotion Recognition
	1 Introduction
	2 Related Work
	3 Dataset
		3.1 Data Preprocessing
		3.2 Feature Extraction
	4 Neural Network Architectures
		4.1 Audio Models
		4.2 Text Models
	5 Experiments
		5.1 Evaluation Metrics
		5.2 Results
	6 Conclusion and Future Work
	References
A Comprehensive Review on Analysis Methods, Software, and Application of fMRI for Classification of Alzheimer’s Disease
	1 Introduction
	2 Functional MRI
	3 Review of Major fMRI Software Projects
	4 State of the Art: Uses of fMRI for Classification and Prediction
	5 Conclusion and Future Work
	References
A Pilot Study of Indian Ragas-Based Music Therapy for Enhancement of Chronic Patient Health Condition
	1 Introduction
		1.1 Healing Diseases with the Music Therapy
		1.2 Ragas and Emotions
		1.3 Raga and Time
	2 Related Works
	3 Indian Raga-Based Music Therapy Approach
		3.1 Blood Pressure Measurement Module
		3.2 Electroencephalography (EEG)
		3.3 Stress Reduction Using Raga Music Therapy
		3.4 Diabetes BSL Measurement Module
	4 Mobile App Development Process Flow
	5 Discussion
	6 Conclusion
	References
Scottish Education: A Model for Strengthening Mental Health—A Case Study in Inclusive Scottish Primary Schools
	1 Introduction
	2 Conceptual Study: Benefits of Music to Mental Health
	3 Empirical Research: Mental Health in Educational Practice—Research Objectives, Methodologies, and Strategies
		3.1 Research Objectives
		3.2 Research Methodologies and Strategies
	4 Research Places and Participants
	5 Results: How Schools Are Striving to Improve Mental Health
	6 Analysis-Eminent Features and Theme: How Schools Are Striving to Improve Mental Health
	7 Conclusion
	References
Spoken Language Processing
Language Model Fine-Tuning with Second-Order Optimizer
	1 Introduction
	2 Experimental Setup
		2.1 Model Architecture
		2.2 Dataset
	3 Methodology
		3.1 Quasi-Newton Template
	4 Results and Analysis
		4.1 CustomModel on MNIST LBFGS Versus Adam
		4.2 ResNeT50 on Cactus Dataset LBFGS Versus Adam
		4.3 MobileNetV2 on CACTUS Dataset LBFGS Versus Adam
		4.4 BERT on SQuAD
	5 Conclusion
	6 Future Work
	References
Human-to-Robotic Voice Conversion by Employing Wave Patterns Based on Gender and Emotions
	1 Introduction
	2 Related Work
	3 Dataset
	4 System 1: Emotion-Gender Classification
		4.1 Module 1: Emotion Identifier
		4.2 Module 2: Gender Predictor
	5 System 2: Robotic Voice Conversion
		5.1 Human Voice Wave Repository
		5.2 Robotic Voice Wave Repository
		5.3 Phoneme Extractor
		5.4 Wave Matching Implementation
		5.5 Wave Replacer
		5.6 Wave Concatenation
	6 Results
	7 Conclusions
	References
Classifying Hate Speeches Shared in Twitter
	1 Introduction
	2 Related Work
	3 System Design
		3.1 Feature Vector Preparation
	4 Experiments and Results
		4.1 Module 1: Ideal Case with All Unique Words
		4.2 Module 2: 4-Gram
		4.3 Module 3: 4-Gram with Sentiment
		4.4 Module 4: Only Sentiment
		4.5 Module 5: Sentence2Vector
		4.6 Case Studies on Module 3
	5 Error Analysis and Observations
	6 Conclusion and Future Work
	References
Impact of Visual Representation of Audio Signals for Indian Language Identification
	1 Introduction
	2 Related Works
	3 Proposed Methodology
		3.1 Mel-Spectrogram Representation
		3.2 Deep Feature Extraction Using CNN
		3.3 List of Comparative Algorithms
	4 Results and Discussion
		4.1 Evaluation Metrics
	5 Conclusion
	References
Bi-Lingual TDNN-LSTM Acoustic Modeling for Limited Resource Hindi and Marathi Language ASR
	1 Introduction
	2 Acoustic Models Exploiting Variable-Length Contextual Information
		2.1 Sub-sampled TDNN
		2.2 Long Short-Term Memory (LSTM) Model
		2.3 TDNN-LSTM Model
	3 Multilingual Architecture for Hindi and Marathi Language
	4 Corpus Details
	5 Experimental Setup
		5.1 Language Modeling
		5.2 Acoustic Modeling
		5.3 Pronunciation Lexicon
		5.4 Tools and Performance Measures
	6 Experimental Results and Discussion
		6.1 Training with One Language
		6.2 Training in Two Languages
	7 Conclusion
	References
Vowel-Based Acoustic and Prosodic Study of Three Manipuri Dialects
	1 Introduction
	2 The Dialect Dataset
		2.1 Annotation of Data
		2.2 Vowel Frequency Distribution
	3 Feature Extraction and Preliminary Analysis
		3.1 Formant Analysis
		3.2 Segment Duration
		3.3 Energy Analysis
		3.4 Pitch Analysis
	4 Conclusion and Future Work
	References
A Continuous Speech Recognition System for Bangla Language
	1 Introduction
		1.1 Bangla Language
		1.2 Related Research Work
	2 Spoken Language Resources
		2.1 Text Corpus
		2.2 Speech Data
		2.3 Pronunciation Dictionary
	3 Experimental Results and Discussion
		3.1 Performance Evaluation Criterion
		3.2 Baseline ASR System
		3.3 Effect of Language Model
		3.4 Optimal Number of Senones
		3.5 Threefold Cross-Validation
		3.6 Language Model Training
	4 Conclusions
	References
Indian Language Speech Label (ILSL): A De Facto National Standard
	1 Introduction
		1.1 International Scene
		1.2 National Scene
	2 Indian Language Speech Labels
		2.1 Vowels
		2.2 Consonants
		2.3 Language-Specific Notes
		2.4 Discussion
	3 Conclusions
	References
Author Index