PRICAI 2023: Trends in Artificial Intelligence: 20th Pacific Rim International Conference on Artificial Intelligence, PRICAI 2023, Jakarta, Indonesia, ... (Lecture Notes in Artificial Intelligence)

Preface
Organization
Contents – Part III
Vision and Perception
A Multi-scale Densely Connected and Feature Aggregation Network for Hyperspectral Image Classification
	1 Introduction
	2 Proposed Method
		2.1 Spectral-Spatial Feature Extraction Module
		2.2 Multi-scale Feature Extraction Module
		2.3 Multi-level Feature Aggregation Module
	3 Experiment and Analysis
		3.1 Dataset Description and Experiment Setup
		3.2 Experiment Results and Analysis
		3.3 Parametric Analysis
		3.4 Ablation Experiments
	4 Conclusion
	References
A-ESRGAN: Training Real-World Blind Super-Resolution with Attention U-Net Discriminators
	1 Introduction and Motivation
	2 Related Work
		2.1 GANs-Based Blind SR Methods
		2.2 Discriminator Models
	3 Method
	4 Experiments
		4.1 Implementation Details
		4.2 Testsets and Experiment Settings
		4.3 Comparing with the State-of-the-Arts
		4.4 Attention Block Analysis
		4.5 Multi-scale Discriminator Analysis
		4.6 Ablation Study
	5 Conclusions
	References
AI-Based Intelligent-Annotation Algorithm for Medical Segmentation from Ultrasound Data
	1 Introduction
		1.1 Contributions
		1.2 Related Work
	2 Methodology
		2.1 Workflow
		2.2 Adaptive Polygon Tracking (APT) Model
		2.3 Historical Storage-Based Quantum-Inspired Evolutionary Network (HQIE)
		2.4 Mathematical Model-Based Contour Detection
	3 Experiment Setup and Results
		3.1 Databases
		3.2 Performance on the Testing Dataset Disturbed by Noise
		3.3 Ablation Study
		3.4 Comparison with State-Of-The-Art (SOTA) Models
	4 Conclusion
	References
An Automatic Fabric Defect Detector Using an Efficient Multi-scale Network
	1 Introduction
	2 Related Work
	3 Proposed Model EMSD
		3.1 LSC-Darknet
		3.2 DCSPPF
		3.3 LSG-PAFPN
		3.4 Detection Head
	4 Experiments
		4.1 Setup
		4.2 Datasets
		4.3 Evaluation Metrics
		4.4 Comparison Experiment Results
		4.5 Ablation Experiments
		4.6 Visualization of Detection Results
	5 Conclusion
	References
An Improved Framework for Pedestrian Tracking and Counting Based on DeepSORT
	1 Introduction
	2 FR-DeepSort for Pedestrian Tracking and Counting
		2.1 The FR-DeepSORT Framework
		2.2 Pedestrian Tracking
		2.3 Pedestrian Counting
	3 Experiments
		3.1 Analysis of Pedestrian Tracking Results
		3.2 Analysis of Pedestrian Counting Results
	4 Conclusion
	References
Bootstrap Diffusion Model Curve Estimation for High Resolution Low-Light Image Enhancement
	1 Introduction
	2 Related Work
		2.1 Learning-Based Methods in LLIE
		2.2 Diffusion Models
	3 Methodology
		3.1 Curve Estimation for High Resolution Image
		3.2 Bootstrap Diffusion Model for Better Curve Estimation
		3.3 Denoising Module for Real Low-Light Image
	4 Experiments
		4.1 Datasets Settings
		4.2 Comparison with SOTA Methods on Paired Data
		4.3 Comparison with SOTA Methods on Unpaired Data
		4.4 Ablation Study
	5 Conclusion and Limitation
	References
CoalUMLP: Slice and Dice! A Fast, MLP-Like 3D Medical Image Segmentation Network
	1 Introduction
	2 Method
		2.1 Overview
		2.2 Multi-scale Axial Permute Encoder
		2.3 Masked Axial Permute Decoder
		2.4 Semantic Bridging Connections
	3 Experiment
		3.1 Dataset
		3.2 Implement Details
		3.3 Comparison with SOTA
		3.4 Ablation Study
	4 Conclusion
	References
Enhancing Interpretability in CT Reconstruction Using Tomographic Domain Transform with Self-supervision
	1 Introduction
	2 Methodology
		2.1 Radon Transform in CT Imaging
		2.2 CT Reconstruction Using Tomographic Domain Transform with Self-supervision
	3 Experimental Results
		3.1 Datasets and Experimental Settings
		3.2 Comparison Experiments
	4 Conclusion
	References
Feature Aggregation Network for Building Extraction from High-Resolution Remote Sensing Images
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Transformer Encoder
		3.2 Feature Aggregation Module
		3.3 Feature Refinement via Difference Elimination Module and Receptive Field Block
		3.4 Dual Attention Module for Enhanced Feature Interactions
		3.5 Fusion Decoder and Loss Function
	4 Experiments
		4.1 Datasets
		4.2 Implementation Details
		4.3 Comparison with Other State-of-the-Art Methods
		4.4 Ablation Study
	5 Conclusion
	References
Image Quality Assessment Method Based on Cross-Modal
	1 Introduction
	2 Related Work
		2.1 Deep Learning-Based Image Quality Assessment
		2.2 Cross-Modal Techniques
	3 Methods
		3.1 Exploring the Feasibility of Cross-Modal Models
		3.2 Image Quality Score Assessment Based on Cross-Modality
	4 Experiments
		4.1 Datasets
		4.2 Experimental Details
		4.3 Evaluation Metrics
		4.4 Feasibility Research
		4.5 Comparison Experiments
		4.6 Ablation Experiments
	5 Conclusion
	6 Outlook
	References
KDED: A Knowledge Distillation Based Edge Detector
	1 Introduction
	2 Related Work
		2.1 Label Problems in Edge Detection
		2.2 Knowledge Distillation
	3 Method
		3.1 Compact Twice Fusion Network for Edge Detection
		3.2 Knowledge Distillation Based on Label Correction
		3.3 Sample Balance Loss
	4 Experiments
		4.1 Datasets and Implementation
		4.2 Comparison with the State-of-the-Art Methods
		4.3 Ablation Study
	5 Conclusion
	References
Multiple Attention Network for Facial Expression Recognition
	1 Introduction
	2 Related Work
		2.1 Real-Time Classification Networks
		2.2 Attention Mechanism
	3 Methodology
		3.1 Multi-branch Stack Residual Network
		3.2 Transitional Attention Network
		3.3 Appropriate Cascade Structure
	4 Experiments
		4.1 Implementation Details
		4.2 Ablation Studies
		4.3 Comparision with Previous Results
	5 Conclusion
	References
PMT-IQA: Progressive Multi-task Learning for Blind Image Quality Assessment
	1 Introduction
	2 Related Works
	3 Methods
		3.1 Overview of the Proposed Model
		3.2 Multi-scale Semantic Feature Extraction
		3.3 Progressive Multi-Task Image Quality Assessment
	4 Experiment
		4.1 Experimental Setup
		4.2 Performance Evaluation
		4.3 Ablation Study
	5 Conclusion
	References
Reduced-Resolution Head for Object Detection
	1 Introduction
	2 Related Works
	3 Method
		3.1 Motivation and Analysis
		3.2 Reduced-Resolution Head for Object Detection
	4 Experiments
		4.1 Ablation Study
		4.2 Applied to Other Detectors
	5 Conclusion
	References
Research of Highway Vehicle Inspection Based on Improved YOLOv5
	1 Introduction
	2 Related Work
		2.1 YOLOv5 Model
		2.2 The Improvement of YOLOv5
	3 Method
		3.1 Ghostnet-C
		3.2 GSConv+Slim-Neck
		3.3 CAS Attention Mechanism
	4 Experiment and Metrics
		4.1 Experimental Environment and Data Set
		4.2 Metrics
		4.3 Experiment and Experimental Analysis
	5 Conclusion
	References
STN-BA: Weakly-Supervised Few-Shot Temporal Action Localization
	1 Introduction
	2 Related Work
	3 Method
		3.1 Feature Extractor
		3.2 Similarity Generator
		3.3 Video-Level Classifier
		3.4 Localization and Boundary-Check Algorithm
	4 Experiment
		4.1 Experiment Setup
		4.2 Main Experimental Results
		4.3 Ablation Experiment
		4.4 Generalization Test
	5 Conclusion
	References
SVFNeXt: Sparse Voxel Fusion for LiDAR-Based 3D Object Detection
	1 Introduction
	2 Related Work
		2.1 Voxel-Based 3D Detectors
		2.2 Fusion-Based 3D Detectors
		2.3 Transformer-Based 3D Detectors
	3 SVFNeXt for 3D Object Detection
		3.1 Dynamic Distance-Aware Cylindrical Voxelization
		3.2 Foreground Centroid-Voxel Selection-Query-Fusion
		3.3 Object-Aware Center-Voxel Transformer
		3.4 Loss Functions
	4 Experiments
		4.1 Datasets
		4.2 Implementation Details
		4.3 Main Results
		4.4 Ablation Study
	5 Conclusion
	References
Traffic Sign Recognition Model Based on Small Object Detection
	1 Introduction
	2 Related Work
		2.1 Data Augmentation
		2.2 Loss Function
		2.3 Deep Learning For Small Object Detection
	3 Method
		3.1 FlexCut Data Augmentation
		3.2 Keypoint-Based PIoU Loss Function
		3.3 The Proposed YOLOv5T
	4 Experiments
		4.1 Dataset
		4.2 Experimental Analysis
	5 Conclusion
	References
A Multi-scale Multi-modal Multi-dimension Joint Transformer for Two-Stream Action Classification
	1 Introduction
	2 The Proposed Method
		2.1 Training Schemes
	3 Experiments
		3.1 Experimental Setups
		3.2 Results and Discussions
		3.3 Visualizations
	4 Conclusions
	References
Adv-Triplet Loss for Sparse Attack on Facial Expression Recognition
	1 Introduction
	2 Method
		2.1 Problem Definition
		2.2 Adv-Triplet Loss Function
		2.3 Adv-Triplet Loss Search Attack
	3 Experiments and Results
		3.1 Sparsity Evaluation
		3.2 Invisibility Evaluation
	4 Conclusion
	References
Credible Dual-X Modality Learning for Visible and Infrared Person Re-Identification
	1 Introduction
	2 Methodology
		2.1 Overview
		2.2 Dual-X Module
		2.3 Uncertainty Estimation Algorithm
	3 Experiment and Analysis
		3.1 Experimental Settings
		3.2 Ablation Study
		3.3 Comparison with State-of-the-Art Methods
	4 Conclusion
	References
Facial Expression Recognition in Online Course Using Light-Weight Vision Transformer via Knowledge Distillation
	1 Introduction
	2 Related Work
	3 Method
	4 Experiments Results
	5 Conclusion
	References
AI Impact
A Deep Reinforcement Learning Based Facilitation Agent for Consensus Building Among Multi-Round Discussions
	1 Introduction
	2 Related Work
	3 Problem Formulation
		3.1 Problem Description
		3.2 Formulation of Multi-round Discussion as a MDP
	4 Algorithm
	5 Evaluation
		5.1 Evaluation Setting
		5.2 Evaluation Results
	6 Conclusion
	References
A Heuristic Framework for Personalized Route Recommendation Based on Convolutional Neural Networks
	1 Introduction
	2 Basic Concepts of PRR
	3 PNNH Framework
		3.1 Design of PNNH Framework
		3.2 Evaluation of PNNH Framework
	4 NDA* Algorithm Based on PNNH Framework
		4.1 Preference Modeling Based on NeuroMLR Neural Network
		4.2 Route Heuristic Algorithm
	5 Experimental Validation and Result Analysis
		5.1 Experimental Setup
		5.2 Baselines
		5.3 Experimental Results and Analysis
	6 Conclusion
	References
Approximate Supplement-Based Neighborhood Rough Set Model in Incomplete Hybrid Information Systems
	1 Introduction
	2 Preliminaries
		2.1 Incomplete Hybrid Information Systems (IHISs)
	3 Approximate Supplement-Based NRSM
		3.1 Approximate Supplement in IHIS
		3.2 Construction of AS-NRSM in IHIS*
	4 Experiments and Analysis
		4.1 Performance Comparisons of Different Algorithms
	5 Conclusion and Future Work
	References
Attention-Guided Self-supervised Framework for Facial Emotion Recognition
	1 Introduction
	2 Methodology
		2.1 CARD Model
		2.2 Convolutional Block Attention Modules (CBAM)
		2.3 Simple Contrastive Learning (SimCLR)
		2.4 Architecture Details
	3 Experiments
		3.1 FER2013 Dataset
		3.2 Performance Evaluation Metrics
		3.3 Implementation Details
	4 Results
	5 Conclusion
	References
BeECD: Belief-Aware Echo Chamber Detection over Twitter Stream
	1 Introduction
	2 Related Work
		2.1 Echo Chambers on Social Platforms
		2.2 Content-Based Echo Chamber Detection
	3 Belief-Aware Echo Chamber Detection
		3.1 Formal Definitions
		3.2 Belief Graph Construction
		3.3 Belief Graph Partitioning
		3.4 Echo Chamber Detection
	4 Experiments and Analysis
		4.1 Data Collection and Organisation
		4.2 Experiment 1: Response Analysis
		4.3 Experiment 2: Belief Graph Impact Analysis
	5 Conclusion and Prospective Research Directions
	References
Building an Egyptian-Arabic Speech Corpus for Emotion Analysis Using Deep Learning
	1 Introduction
	2 Related Work
	3 Dataset Collection
	4 Dataset Description
	5 Dataset Validation
	6 Deep Learning Model
	7 Independent-Speaker Multi-class Emotion Classification Results
	8 Conclusion
	References
Finding the Determinants of Lower Limb Amputations Related to Diabetic Foot Ulcer - A Logistic Regression Classifier
	1 Introduction
	2 Literature Review
	3 Research Methodology
		3.1 Data Collection
		3.2 Classification Model
	4 Results and Discussion for the Binary Logistic Regression Classifier
		4.1 Model Evaluation
	5 Conclusion
	References
Frequency Domain Feature Learning with Wavelet Transform for Image Translation
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Wavelet Transform Based Image Processing
		3.2 Training Objectives
	4 Experiments
		4.1 Baselines
		4.2 Dataset and Evaluation Indicators
		4.3 Implementation Details
		4.4 Experimental Results
		4.5 Ablation Study
	5 Conclusion
	References
Graph-Guided Latent Variable Target Inference for Mitigating Concept Drift in Time Series Forecasting
	1 Introduction
	2 Related Work
	3 Graph-Guided Latent Variable Target Inference
		3.1 Variable Target Inference
		3.2 General Architecture
		3.3 The Evolution of Latent Graphs
		3.4 Decoding and Forecasting
	4 Experiments
		4.1 Evaluating Predictive Performance
		4.2 Ablation Experiments
	5 Conclusion
	References
Optimization of Takagi-Sugeno-Kang Fuzzy Model Based on Differential Evolution with Lévy Flight
	1 Introduction
	2 TSK Fuzzy Neural Network
		2.1 Classical TSK Fuzzy Model
		2.2 Optimization Techniques Implementation
	3 DELF Algorithm Design
		3.1 Optimization Task Formulation
		3.2 Evaluation Function Design
		3.3 Mutation Operator with Lévy Flight
		3.4 Crossover Operator
	4 Simulation Result Analysis
		4.1 Searching Ability Comparison
		4.2 TSK Fuzzy Model Optimization
	5 Conclusion
	References
RPL-SVM: Making SVM Robust Against Missing Values and Partial Labels
	1 Introduction
	2 Multiclass Partial Label SVM (PL-SVM)
	3 Imputing Missing Values
	4 RPL-SVM Linear: Robust Formulation for Linear Classifiers with Partial Labels and Missing Values
	5 RPL-SVM Nonlinear: Robust Formulation for Nonlinear Classifiers with Partial Labels
	6 Experiments
		6.1 Experimental Setup
		6.2 Performance Comparison Results of RPL-SVM with Baselines
		6.3 Effect of  on RPL-SVM
	7 Conclusions and Future Work
	References
Spatial Gene Expression Prediction Using Coarse and Fine Attention Network
	1 Introduction
	2 Related Work
	3 Method
	4 Experiments
		4.1 Datasets
		4.2 Experimental Set-Up
		4.3 Experimental Results
		4.4 Ablation Study
	5 Conclusion
	References
STFM: Enhancing Autism Spectrum Disorder Classification Through Ensemble Learning-Based Fusion of Temporal and Spatial fMRI Patterns
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 The Overview of Our Method
		3.2 Brain Network Construction
		3.3 Extraction of Temporal and Spatial Patterns
		3.4 Fusion of Patterns
		3.5 Classifier
	4 Experiments
	5 Conclusion and Future Work
	References
Unified Counterfactual Explanation Framework for Black-Box Models
	1 Introduction
	2 Method
		2.1 UNICE Framework
		2.2 UNICE Implementation
	3 Experiments
		3.1 Experimental Settings
		3.2 UNICE Performance
		3.3 Analysis and Discussion
	4 Conclusion
	References
VIFST: Video Inpainting Localization Using Multi-view Spatial-Frequency Traces
	1 Introduction
	2 Related Work
		2.1 Object Inpainting Localization
		2.2 Vision Transformer
	3 Method
		3.1 Spatial Branch
		3.2 Frequency Branch
		3.3 Learning Local Features
		3.4 Learning Global Contextual Correlation Features
	4 Experiment
		4.1 Experimental Setup
		4.2 Comparison Experiments
		4.3 Robustness Experiments
	5 Ablation Study
		5.1 Influence of Spatial and Frequency Branches
		5.2 Impact of Different Component Combinations
	6 Results Analysis and Discussion
	7 Conclusion
	References
A Logistic Regression Classification Model to Predict ERP Systems Adoption by SMEs in Fiji
	1 Introduction
	2 Literature Review
	3 Research Methodology
	4 Results and Discussion
		4.1 Model Evaluation
	5 Conclusion
	References
Low-Resource Machine Translation Training Curriculum Fit for Low-Resource Languages
	1 Introduction
	2 Related Work
	3 Proposed Methods
		3.1 Dictionary Creation and Sentence Mining
		3.2 Training Curriculum
	4 Experiment Setup and Results
		4.1 Experimental Setup
		4.2 Results
	5 Conclusion
	References
MARL4DRP: Benchmarking Cooperative Multi-agent Reinforcement Learning Algorithms for Drone Routing Problems
	1 Introduction
	2 Drone Routing Problem
		2.1 Definition of the DRP
		2.2 Formulating DRP as a MAPF
	3 Cooperative MARL for DRP
	4 Evaluation
	5 Conclusion
	References
Real-Time Event Detection with Random Forests and Temporal Convolutional Networks for More Sustainable Petroleum Industry
	1 Introduction
	2 Data and Methodology
	3 Evaluation and Conclusion
	References
SLSNet: Weakly-Supervised Skin Lesion Segmentation Network with Self-attentions
	1 Introduction
	2 Lesion Expansion Network with Self-attentions
		2.1 Intra-image Self-attention Seed Expansion (ISE) Module
		2.2 Inter-image Affinity-Based Noise Suppression (IAS) Module
	3 Experiments
		3.1 Dataset and Evaluation Metrics
		3.2 Comparison with Other Methods on the ISIC-2017 Dataset
	4 Conclusion
	5 Compliance with Ethical Standards
	References
Trust and Reputation Management in IoT Using Multi-agent System Approach
	1 Introduction
	2 Related Work
	3 IoT-CADM
		3.1 IoT-CADM Trust Evaluation and Selection Model (IoT-TESM)
	4 Experimental Evaluation
	5 Conclusion and Future Work
	References
Author Index