AI-generated Content: First International Conference, AIGC 2023, Shanghai, China, August 25–26, 2023, Revised Selected Papers (Communications in Computer and Information Science, 1946)

This book constitutes the revised selected papers of the First International Conference, AIGC 2023, held in Shanghai, China, during August 25–26, 2023 The 30 full papers included in this volume were carefully reviewed and selected from 62 submissions. The volume focuses on the remarkable strides that have been made in the realm of artificial intelligence and its transformative impact on content creation. As delving into the content of the proceedings, the readers will encounter cutting-edge research findings, innovative applications, and thought-provoking insights that underscore the transformative potential of AI-generated content. Preface Organization Contents Automatic Multilingual Question Generation for Health Data Using LLMs 1 Introduction 2 Methods 2.1 Data Acquisition 2.2 Data Preparation 2.3 Language Models 2.4 Experimentation 3 Results 4 Evaluation Metrics 5 Discussion 5.1 RQ1: What Existing Systems or Models can be Used to Automatically Generate Different Sets of Questions Given the Context for English as Well as Spanish? 5.2 RQ2: Is the Quality of Questions Generated Consistent Across Different Texts and Thresholds? If the Quality is not Consistent, then What Factors Impact or Determine the Quality of Questions Generated? 5.3 RQ3: What are the Optimal Values for These Factors Impacting the Question Generation Quality? 5.4 Evaluation References DSEN: A Distance-Based Semantic Enhancement Neural Network for Pronoun Anaphora Resolution 1 Introduction 2 Related Work 3 Model 3.1 Word Embedding 3.2 Distance-Embedding 3.3 Self-attention 3.4 Fine-Grained Scoring 4 Experiments 4.1 Datasets 4.2 Parameter Settings 4.3 Comparsion with Other Mehods 4.4 Ablation Studies 5 Conclusions References Inheritance and Revitalization: Exploring the Synergy Between AIGC Technologies and Chinese Traditional Culture 1 Introduction 2 Related Work 2.1 Text-to-Image Synthesis 2.2 Large Language Models 2.3 LoRA 3 Method 3.1 Prompt Building 3.2 Finetuning with LoRA 3.3 Workflow of the Inference Pipeline 4 Experiments 4.1 Settings 4.2 Results 5 Conclusions References Strong-AI Autoepistemic Robots Build on Intensional First Order Logic 1 Introduction 2 Algebra for Composition of Meanings in IFOL 3 Four-Levels Robot's Brain Structure 4 A Short Introduction to Robots Autoepistemic Deduction 5 Conclusions and Future Work References Divide and Control: Generation of Multiple Component Comic Illustrations with Diffusion Models Based on Regression 1 Introduction 2 Related Work 2.1 Diffusion Probabilistic Models 2.2 Controllable Text-to-Image Generation with Diffusion Models 3 Method 3.1 Component Image Generation 3.2 Regression 3.3 Composition 3.4 Final Image Generation 4 Experiments 5 Conclusions References Enhancing EFL Vocabulary Acquisition Through Computational Thinking 1 Introduction 1.1 ChatGPT and its Essence 1.2 Thinking of AI and LI 1.3 Computational Thinking (CT) 1.4 Research Questions 2 Methods 2.1 Quasi-experimental Design 2.2 Experimental Conditions 2.3 Participants 2.4 Assessment Tools 2.5 Intervention Process 2.6 Data Collection and Processing 3 Results 3.1 Comparison Between the Experimental Class and Control Class 3.2 The Experimental Class 3.3 The Control Class 4 Discussion 4.1 How Does Computational Thinking Intervene in English Vocabulary Acquisition? 4.2 How Does Computational Thinking Affect English Vocabulary Acquisition? 4.3 What is the Role of Computational Thinking in Vocabulary Intervention? 5 Conclusion References Connected-CF2: Learning the Explainability of Graph Neural Network on Counterfactual and Factual Reasoning via Connected Component 1 Introduction 2 Preliminaries and Notation 2.1 Learning Representations 2.2 Graph Classification and Node Classification 2.3 Laplacian Matrix and Connected Component 3 Problem Formulation 3.1 Explainable Graph Neural Networks 3.2 Reasoning Conditions 3.3 Compact, Effective and Connected Explanations 4 The Connected-CF2 Framework 4.1 Connected-CF2 Optimization Problem 4.2 Relaxation Optimization 5 Experiments 5.1 Dataset 5.2 Baselines 5.3 Setup 5.4 Quantitative Analysis 5.5 Qualitative Analysis 6 Conclusion and Future Work References Controllable Feature-Preserving Style Transfer 1 Introduction 2 Related Work 2.1 StyleGAN-Related Models 2.2 Evaluation of Style Transfer 2.3 Developments of Style Transfer 3 Method 3.1 Image Style Transfer 3.2 Style Transfer Evaluation 3.3 Aesthetics Evaluation 4 Experiments 4.1 Datasets 4.2 Implementation Details 4.3 Comparable Experiments 5 Conclusion References Data Adaptive Semantic Communication Systems for Intelligent Tasks and Image Transmission 1 Introduction 2 System Model 3 Proposed Semantic Communication Networks 3.1 Network Structures of Encoder and Decoder 3.2 Network Structures of Classifier 3.3 Model Training 4 Experimental Results 4.1 Image Recovery 4.2 Data Adaptation 4.3 Classification Task 5 Conclusion References Generative Artificial Intelligence and Metaverse: Future of Work, Future of Society, and Future of Humanity 1 Introduction: Generative Artificial Intelligence and Metaverse 1.1 Generative Artificial Intelligence 1.2 Metaverse 2 Current Development of GenAI and Metaverse 2.1 Practical Applications 2.2 Research Topics 3 Potential of Integrating GenAI into Metaverse 3.1 Harnessing Rich Data in Metaverse 3.2 Capitalizing the Dynamic Nature of Metaverse 3.3 Enhancing Ubiquitous Human-AI Interactions in Metaverse 3.4 Enabling Collaborative Creation in Metaverse 4 Research Directions and Research Questions 4.1 Future of Work 4.2 Future of Society 4.3 Future of Humanity 4.4 Integrating GenAI and Metaverse 5 Conclusions References CanFuUI: A Canvas-Centric Web User Interface for Iterative Image Generation with Diffusion Models and ControlNet 1 Introduction 2 Related Work 3 Canvas-Centric Web UI 3.1 AI Artistry 3.2 Canvas and Toolbar 3.3 Asset Library 3.4 Use Case 4 Comparison with WebUI and ComfyUI 5 Implementation 6 Evaluation 7 Conclusion References MFAR-VTON: Multi-scale Fabric Adaptive Registration for Image-Based Virtual Try-On 1 Introduction 2 Related Work 2.1 Trainable Image Alignment 2.2 Image-Based Virtual Try-On 3 Proposed Method 3.1 Segmentation Generation Module 3.2 Multi-scale Neighborhood Consensus Warp Module 3.3 Try-On Generator Module 4 Experiments 4.1 Dataset 4.2 Implementation Details 4.3 Qualitative Results 4.4 Quantitative Results 4.5 Ablation Study 5 Conclusion References An Assessment of ChatGPT on Log Data 1 Introduction 2 Related Work 2.1 Log Data 2.2 Log Processing 2.3 ChatGPT 2.4 ChatGPT Evaluation 3 Context 3.1 Research Questions 3.2 Dataset 3.3 Experimental Setup 3.4 Evaluation Metrics 4 Experiments and Results 4.1 Log Parsing 4.2 Log Analytics 4.3 Log Summarization 5 Discussion 5.1 Handling Unstructured Log Data 5.2 Performance with Zero-Shot Learning 5.3 Scalability - Message Cap for GPT 5.4 Latency 5.5 Privacy 6 Conclusion 7 Disclaimer References DSQA-LLM: Domain-Specific Intelligent Question Answering Based on Large Language Model 1 Introduction 2 Related Work 2.1 Question Classification 2.2 Question Embedding 2.3 Answer Extraction and Generation 3 Approach 3.1 Overview 3.2 Question Processing Phase 3.3 Similarity Searching Phase 4 Experiment 4.1 Datasets 4.2 Metrics and Baseline 4.3 Results 5 Conclusion References Research on the Construction of Intelligent Customs Clearance Information System for Cross-Border Road Cargo Between Guangdong and Hong Kong 1 Introduction 2 Scheme of Intelligent Customs Clearance Information System between GD and HK 3 The Implementation of AI Machine Learning 4 The Key Applications of AI 4.1 Intelligent Classification of Hong Kong Commodities Based on Machine Translation and Text Similarity 4.2 Intelligent Recognition of Customs Documents and Generation of Customs Declaration Draft in GD and HK 4.3 Intelligent Extraction of Commodity Declaration Element Information 5 Conclusion References Analyzing Multilingual Automatic Speech Recognition Systems Performance 1 Introduction 2 Open-Source ASR Systems 2.1 Whisper 2.2 Vosk 2.3 Kaldi 2.4 Julius 2.5 Mozilla DeepSpeech 3 Methods 3.1 Data 3.2 Data Preprocessing 3.3 Human Transcription 3.4 Automatic Transcription 3.5 Experiments 4 Results 4.1 Transcription Models Performance 5 Transcription Evaluation 5.1 ASR Evaluation 5.2 ASR Transcriptions Error Rates 5.3 WER Related Challenges 6 Discussion 6.1 RQ1: What Open-Source ASR Systems Exist that can Transcribe English as well as Spanish Videos? 6.2 RQ2: What Models within these Systems can be Used to Generate Transcriptions for Recorded Videos based on the Performance of the Models/systems for the Two Languages? 6.3 RQ3: What Evaluation Measures can be Used to Automatically Evaluate the System/model’s Performance? 6.4 Limitations 7 Conclusions References Exploring User Acceptance of Al Image Generator: Unveiling Influential Factors in Embracing an Artistic AIGC Software 1 Introduction 2 Literature Review 2.1 AI-based Art Content Generation and Relevant Theoretical Frameworks 2.2 AI-based Art Content Generation and Relevant Theoretical Frameworks 2.3 Characteristics of Al Image Generator 3 Research Methodology 4 Data Analysis and Results 4.1 AI-based Art Content Generation and Relevant Theoretical Frameworks 4.2 Skewness and Kurtosis Statistical Analysis 4.3 Variable Correlation Analysis 4.4 Variable Correlation Analysis 5 Conclusions References File Management System Based on Intelligent Technology 1 The Introduction 2 Overview of Artificial Intelligence and Archive Management 2.1 Artificial Intelligence Technology 2.2 Intelligent Management of Archives Under Artificial Intelligence Technology 2.3 The Impact of Artificial Intelligence on Archive Management 3 Design and Process of Intelligent Models for File Management 3.1 System Architecture 3.2 Process Flow 4 File Intelligent Management Model Implementation 4.1 The Efficiency of File Identification and Collation Has Been Significantly Improved 4.2 The Results of File Identification and Collation Are More Standardized 4.3 The File Identification and Collation Process is More Optimized 5 Conclusion References Leveraging Machine Learning for Crime Intent Detection in Social Media Posts 1 Introduction 2 Methods 2.1 About the Dataset 2.2 Data Preprocessing 2.3 Feature Extraction 2.4 Tokenization 2.5 Dataset Splitting 2.6 Model Training 3 Results 4 Discussion 5 Conclusion References Generation of Visualized Medical Rehabilitation Exercise Prescriptions with Diffusion Models 1 Introduction 2 Related Work 2.1 Medical Prescription Visualization 2.2 Diffusion Models and ControlNet 3 Methods 3.1 Prompt-Only 3.2 Openpose with ControlNet 3.3 Canny with ControlNet 3.4 Scribble with ControlNet 4 Experiments 5 Conclusion and Future Work References A Literature Analysis of the Application of Artificial Intelligence in Judicial Adjudication 1 Introduction 2 Literature Analysis 2.1 Data Collection 2.2 Keywords Analysis 2.3 Analysis of Bibliographic Citation 3 Application of AI in Judicial Adjudication 3.1 The Application of AI in Procedural Matters of Judicial Adjudication 3.2 The Application of AI in Substantial Matters of Judicial Adjudication 3.3 The Controversy of Legal Values Beyond the Application of AI in Legal Practice 4 Conclusion References Overlapping Fingerprint Segmentation and Separation Based on Deep Learning 1 Introduction 2 Related Work 2.1 Fingerprint Image Generation 2.2 Fingerprint Image Segmentation 3 Method 3.1 Data Preprocessing 3.2 Fingerprint Segmentation Model 3.3 Overlapping Fingerprint Separation 4 Experiments and Results 5 Conclusions References Artificial Intelligence for Text Generation: An Intellectual Property Perspective 1 Introduction 2 The Technical Logic of the AI-Generated Text 2.1 General Description 2.2 Specific Illustration 3 The Current Situation and Dilemma of Intellectual Property Regulation on AI-Generated Text 3.1 States’ Domestic Law 3.2 International Law 3.3 Existing Dilemmas 4 Intellectual Property Regulation of AI-Generated Text: Multi-actor Improvement Measures 4.1 Self-censorship 4.2 Self-improvement: Ensuring the Originality of the AI-generated Text 4.3 Promoting International Cooperation: Balancing Data Processing and Protection 4.4 Strengthening Legislation 5 Conclusion References Generative Artificial Intelligence Upgrading Design of TCM Medical Diagnosis System Integration Platform 1 Background 1.1 Pain Points of Medical Technical Diagnosis 1.2 Artificial Intelligence and Large Number According to Technology 2 The Idea of the Generation of Artificial Intelligence Upgrading of the Integrated Platform of Medical Diagnostic System of Traditional Chinese Medicine 2.1 The Generated Artificial Intelligence of Each System is Combined with the Upgrade Design 2.2 Detailed Description of the Generated Artificial Intelligence of Each System 3 Functional Structure Description of Generative Artificial Intelligence Upgrade Design of Traditional Chinese Medicine Medical Diagnosis System Integration Platform 4 Functional Upgrade Contrast 5 Feasibility Discussion on Generative Artificial Intelligence Upgrade Design Plan 6 Prospect References Convolutional Neural Network (CNN) to Reduce Construction Loss in JPEG Compression Caused by Discrete Fourier Transform (DFT) 1 Introduction 2 Methods 2.1 Network Architecture 2.2 Loss Function 3 Training 3.1 Datasets 3.2 Procedure 4 Evaluation 5 Conclusion References Research on Rural Feature Classification Based on Improved U-Net for UAV Remote Sensing Images 1 Introduction 2 Algorithm Principle 2.1 Algorithm Introduction 2.2 Network Structure 3 Experiment and Analysis 3.1 Network Improvement 4 Results 4.1 Model Accuracy Evaluation 4.2 Analysis of Experimental Results 5 Conclusions References Exploring the Intersection of Complex Aesthetics and Generative AI for Promoting Cultural Creativity in Rural China After the Post-pandemic Era 1 Introduction 1.1 Research Background 1.2 Research Gap 1.3 Contribution 2 Structure of the Research 3 Related Works 3.1 The Most Earlier Examples Internationally of the Art Applications for Rural Revitalization 3.2 The Most Earlier Examples Domestically of the Art Applications for Rural Revitalization 3.3 Latest Application Using AIGC Tools to Boost Rural Tourism and for Rural Revitalization 4 Content Generation Workflow 5 Methodology 5.1 Case Study 5.2 Questionnaire 5.3 Semi-structured Interview 6 Key Findings 7 Research Agenda and Limitation 7.1 Research Agenda 7.2 Limitations 8 Conclusion Remark References Fingerprint Image Enhancement Method Based on U-Net Model 1 Introduction 2 Related Work 3 Dataset 4 Research Methods 5 Experiments and Results 6 Conclusions References Applications and Implication of Generative AI in Non-STEM Disciplines in Higher Education 1 Introduction 2 Large Language Model and Generative AI 3 Generative Artificial Intelligence in Education 3.1 Generative AI Aid in Scientific Writing 3.2 The Potential Benefits of Generative Artificial Intelligence in Various of Disciplines in Higher Education 3.3 Large Language Model’s Application on Finance and Economics Disciplines 4 Recommendation 4.1 Ethics and Morality Challenge 4.2 The Scaffolding Function of Generative AI in Non-STEM Disciplines 4.3 Generative AI as Assistive Tool not Autonomous Force 4.4 Generating Better Academic Assessment with AI 4.5 Exercise Prudence with Generative AI's Imaginative Outputs 5 Conclusion References Industry 4.0 Technology-Supported Framework for Sustainable Supply Chain Management in the Textile Industry 1 Introduction 2 Textile Industry and Sustainability Issues 3 Proposed Framework 3.1 Nature of Data Collected at Different Stages of the Supply Chain 3.2 Data Collection, Storage and the Technologies Involved 3.3 Big-Data-Driven SSCM 4 Conclusion References Author Index