Music Intelligence: Second Summit, SOMI 2023, Beijing, China, October 2830, 2023, Revised Selected Papers (Communications in Computer and Information Science, 2007)

This book constitutes the refereed proceedings of the Second Summit on Music Intelligence, SOMI 2023, held in Beijing, China, during October 28–30, 2023.The 10 full papers included in this book were carefully reviewed and selected from 27 submissions. They were organized in topical sections as follows: Music, Neural Science and Music Therapy, Multimodality in Music and Music Generation. Preface Organization Contents Music, Neural Science and Music Therapy Musical Training Changes the Intra- and Inter-network Functional Connectivity 1 Introduction 2 Methods 2.1 Participants 2.2 MRI Data Collection 2.3 Data Preprocessing 2.4 Network Functional Connectivity Construction 2.5 Statistical Analysis 3 Results 3.1 Intra-network Functional Connectivity 3.2 Inter-network Functional Connectivity 4 Discussion 4.1 Intra-network Functional Connectivity 4.2 Inter-network Functional Connectivity 4.3 Limitation 5 Conclusion References Research on the Improvement of Children's Attention Through Binaural Beats Music Therapy in the Context of AI Music Generation 1 Introduction 2 Literature Review 2.1 ADD and HRV 2.2 BBMT Theory 2.3 AI Music Generation 3 Methods 3.1 Experiment Objective 3.2 Experimental Materials 3.3 Participants 3.4 Experimental Procedure 4 Results 4.1 Effect of Music on Children's Attention 4.2 Effect of AI Music on Children's Attention 4.3 Effect of BB-Music on Children's Attention 4.4 Effect of Combining AI Music and BB Technology on Children's Attention 5 Discussion 5.1 Reception to AI Music and BB Treatment 5.2 Challenges Encountered in the Experiment 5.3 Limitations and Future Research Directions 6 Conclusion References Generated Therapeutic Music Based on the ISO Principle 1 Introduction 2 Related Work 2.1 Emotion-Based Music Generation 2.2 Music Dataset with Valence-Arousal Values 3 ABC Notation Dataset Annotated with Valence-Arousal Values 3.1 Methods 3.2 Analysis 4 Music Therapy System 5 Discussion 6 Conclusion References Multimodality in Music Utilizing Quantum Particle Swarm Optimization for Multimodal Fusion of Gestures and Facial Expressions in Ensemble Conducting for Emotional Recognition 1 Introduction 2 Related Work 2.1 Multimodal Emotion Recognition 2.2 Multimodal Fusion Strategies 2.3 Quantum Particle Swarm Optimization 3 FEGE Dataset 3.1 Platform and Setup 3.2 Data Acquisition 3.3 Data Processing 4 Proposed Method 4.1 Main Modules 4.2 Joint Learning 4.3 Quantum Particle Swarm Optimization-Based Parameter Optimization for Joint Learning in Multimodal Data Fusion 5 Experimental Results and Analysis 5.1 Emotion Recognition Results 5.2 Ablation Analysis and Discussion 6 Conclusion References Guqin Jianzi Notation Transcription Based on Language Model 1 Introduction 2 Related Works 3 Dataset 3.1 Jianzi Notation Dataset Construction 3.2 Special Jianzi Encoding 4 Methodology 4.1 Workflow 4.2 Jianzi Deconstruction 4.3 Jianzi Language Model 4.4 Notation Duration Prediction 5 Experiments and Result Analysis 5.1 Experiment Setup 5.2 Comparison Experiment Analysis 5.3 Jianzi Embedding Analysis 5.4 Data Volume Analysis 6 Conclusion References Music Generation Towards the Implementation of an Automatic Composition System for Popular Songs 1 Introduction 2 Background 3 System Overview 4 Generator Modules 4.1 Score Generator 4.2 Performance Generator 4.3 Audio Generator 4.4 Pipeline 5 Implementation 6 Experiment 7 Conclusion and Future Works References CoCoFormer: A Controllable Feature-Rich Polyphonic Music Generation Method 1 Introduction 2 Related Work 3 Proposed Model CoCoFormer 3.1 Data Representation 3.2 CoCoFormer 3.3 Loss Function 4 Experiment 4.1 Dataset 4.2 Experimental Setup 4.3 Ablation Test 4.4 Comparative Test 4.5 Qualitative Experiment 5 Conclusions References C2-MAGIC: Chord-Controllable Multi-track Accompaniment Generation with Interpretability and Creativity 1 Introduction 2 Related Work 3 Method 3.1 Model Overview 3.2 CVAE Input Module 3.3 KL-Annealing Curve 3.4 Hybrid Chord Representation 3.5 Multi-task Loss Ratio 4 Experiments 4.1 Experimental Settings 4.2 Results 4.3 Further Discussion 4.4 Human Evaluation 5 Conclusion References HRPE: Hierarchical Relative Positional Encoding for Transformer-Based Structured Symbolic Music Generation 1 Introduction 2 Related Work 2.1 Structured Symbolic Music Generation 2.2 Positional Information in Transformer 3 Methodology 3.1 Model Overview 3.2 Bar-Level Relative Positional Encoding 3.3 Section-Level Relative Positional Encoding 4 Implementation Details 4.1 Datasets 4.2 Training and Inference Details 5 Experiments and Results 5.1 Quantitative Analysis 5.2 Human Evaluation 5.3 Case Study 6 Conclusion References Chinese Chorales Dataset: A High-Quality Music Dataset for Score Generation 1 Introduction 2 Related Work 2.1 Existing Music Dataset 2.2 Requirements for the Choral Music Dataset 3 Dataset Description and External Annotation Methods 3.1 Dataset Production Process 3.2 Dataset Format 4 Experiments 4.1 Model and Parameter Settings 4.2 Results 5 Conclusion References Author Index