Human brain and artificial intelligence : third International Workshop, HBAI 2022, held in conjunction with IJCAI-ECAI 2022, Vienna, Austria, July 23, 2022, revised selected papers / / Xiaomin Ying, editor
Communications in computer and information science ; ; 1692
Disciplina
006.3
Soggetti
Artificial intelligence
Brain-computer interfaces
Lingua di pubblicazione
Inglese
Formato
Materiale a stampa
Livello bibliografico
Monografia
Nota di bibliografia
Includes bibliographical references and index.
Nota di contenuto
Intro -- Preface -- Organization -- Contents -- AI for Brain Related Data Analysis -- Classification of EEG Signals Based on GA-ELM Optimization Algorithm -- 1 Introduction -- 2 Optimization of Extreme Learning Machine by Genetic Algorithm -- 3 The Experiment Design -- 3.1 Experimental System Framework -- 3.2 Data Acquisition -- 4 The Data Analysis -- 4.1 Preprocessing -- 4.2 Feature Extraction -- 4.3 Genetic Algorithm Optimized Parameter Setting -- 5 Results -- 6 Discussion -- 7 Conclusion -- References -- Delving into Temporal-Spectral Connections in Spike-LFP Decoding by Transformer Networks -- 1 Introduction -- 2 Methods -- 2.1 Temporal Connection Learning with Spikes -- 2.2 Spectral Connection Learning with LFPs -- 2.3 Temporal-Spectral Connection Learning with Spike-LFPs -- 2.4 Task-Related Output Layer -- 3 Experiments and Results -- 3.1 Clinical Dataset -- 3.2 Spike-LFP Fusion Improves Neural Decoding Accuracy -- 3.3 Temporal Connections Improve Robustness to Temporal Shifts -- 3.4 Temporal-Spectral Connections Improve Robustness to Noises -- 4 Conclusion -- A Detail Settings Of Neural Decoders -- B Estimating Movement Conduction Durations With Neuron Responses -- C Robustness To Gaussian Noises -- References -- A Mask Image Recognition Attention Network Supervised by Eye Movement -- 1
Introduction -- 2 Methods -- 2.1 Datasets -- 2.2 The Generation of Gaze Heat Map -- 2.3 Network Architecture -- 3 Results -- 3.1 Eye Movement Heat Map -- 3.2 Network Performance -- 3.3 Network Attention Visualization -- 4 Conclusion -- References -- DFC-SNN: A New Approach for the Recognition of Brain States by Fusing Brain Dynamics and Spiking Neural Network -- 1 Introduction -- 2 Methods -- 2.1 DFC-SNN Framework -- 2.2 Dataset -- 3 Results -- 4 Conclusion -- References.
DSNet: EEG-Based Spatial Convolutional Neural Network for Detecting Major Depressive Disorder -- 1 Introduction -- 2 Materials and Methods -- 2.1 Dataset and Data Preprocessing -- 2.2 The Architecture of DSNet -- 2.3 Baseline Methods -- 2.4 Model Implementation and Experimental Evaluation -- 3 Results and Discuss -- 4 Conclusion -- References -- SE-1DCNN-LSTM: A Deep Learning Framework for EEG-Based Automatic Diagnosis of Major Depressive Disorder and Bipolar Disorder -- 1 Introduction -- 2 Materials and Methods -- 2.1 Data and Preprocessing -- 2.2 1DCNN and LSTM Network -- 2.3 Channel Attention -- 2.4 Evaluation Metrics and Parameters -- 3 Results and Discussion -- 3.1 Comparison with Baseline Method -- 3.2 Ablation Study -- 3.3 Interpretability Analysis of Channel Attention -- 3.4 Effects of Window Size -- 4 Conclusion -- References -- Emotion Recognition from EEG Using All-Convolution Residual Neural Network -- 1 Introduction -- 2 Methods -- 2.1 Pre-processing and Feature Extraction -- 2.2 3D Input Construction -- 2.3 The All-Convolutional Neural Network -- 2.4 Deep Residual Learning Framework -- 3 Experiments -- 3.1 DEAP Dataset -- 3.2 Model Implementation -- 3.3 Parameter Setting -- 4 Experimental Results and Discussion -- 5 Conclusions and Future Works -- References -- Salient Object Detection with Fusion of RGB Image and Eye Tracking Data -- 1 Introduction -- 2 Method -- 2.1 Acquisition of ETSM -- 2.2 Cross-Modal Fusion Module -- 2.3 Improved Cascade Decoder -- 2.4 Optimization Module -- 3 Experimental Setup and Result Analysis -- 3.1 Experimental Details -- 3.2 Evaluation Indicators -- 3.3 Performance Comparison with Other Algorithms -- 4 Conclusion -- References -- Multi-source Domain Adaptation Based on Data Selector with Soft Actor-Critic -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Reinforcement Learning Data Selector.
3.2 Soft Actor-Critic -- 4 Experiments -- 4.1 Datasets and Experiment Settings -- 4.2 Comparison with the Latest Technology -- 4.3 Training Efficiency of DSAC -- 4.4 Source to Single-Target Adaptation -- 5 Conclusion -- References -- Transfer Learning to Decode Brain States Reflecting the Relationship Between Cognitive Tasks -- 1 Introduction -- 2 Related Work -- 2.1 Cognitive Task Relations from Neuroscience Perspective -- 2.2 Cognitive Task Relations From Transfer Learning Perspective -- 3 Methods -- 3.1 HCP Tasks -- 3.2 Transfer Learning -- 3.3 Validation of Cognitive Taskonomy -- 4 Results -- 4.1 Affinity Matrix of Cognitive Tasks -- 4.2 Compare with Task-Specific Networks -- 4.3 Brain Decoding Accuracy with Transfer Learning -- 5 Discussion -- References -- AI and Brain Interface -- Brain Network Analysis of Hand Motor Execution and Imagery Based on Conditional Granger Causality -- 1 Introduction -- 2 Methods -- 2.1 Conditional Granger Causality Analysis -- 3 Data Collection and Processing -- 3.1 Data Collection -- 3.2 Data Processing -- 4 Result -- 5 Discussion -- References -- A Hybrid Brain-Computer Interface for Smart Car Control -- 1 Introduction -- 2 Materials and Methods -- 2.1 System Overview -- 2.2 Signal Acquisition -- 3 Experiments and Results -- 3.1 Experiment I: Single-Mode Control -- 3.2 Experiment II: Multimodal Car Control -- 3.3 Results -- 4 Conclusions -- References -- A Spiking
Neural Network for Brain-Computer Interface of Four Classes Motor Imagery -- 1 Introduction -- 2 Methods -- 2.1 BSA Based on Parameter-Wise Gradient Descent Optimization Method -- 2.2 LIF Model and ALIF Model -- 2.3 The Architecture of SNN -- 2.4 Surrogate Gradient -- 2.5 Channel-Wise Normalization -- 3 Experiments -- 3.1 Dataset and the Selection of Experimental Data -- 3.2 The Performance of PW-GD Optimizing for BSA.
3.3 The Performance of SNN in MI Classification -- 3.4 Comparison of Training Effects of MG and Slayer -- 4 Conclusion -- References -- Virtual Drone Control Using Brain-Computer Interface Based on Motor Imagery Brain Magnetic Fields -- 1 Introduction -- 2 Brain Magentic Fields Based BCI System Description -- 2.1 Helmet Design -- 2.2 Experimental Environment Construction -- 3 Experiment Content and Data Processing -- 3.1 Subject Training -- 3.2 Data Collection and Processing -- 4 Machine Learning and Result -- 4.1 Data for Machine Learning -- 4.2 Model Training -- 4.3 Test Result -- 4.4 Virtual Drone Controlling -- 5 Conclusion and Future Work -- References -- Brain Controlled Manipulator System Based on Improved Target Detection and Augmented Reality Technology -- 1 Introduction -- 2 Experimental and System Structure Design -- 2.1 Experimental Design -- 2.2 System Structure Design -- 3 Methods -- 3.1 Improvement of Faster-RCNN -- 3.2 AR Technology Generates Stimulation Interface -- 3.3 EEG Signal Analysis -- 4 Experimental results -- 4.1 Target Detection Model Test -- 4.2 EEG Recognition Results Using AR -- 4.3 System Test -- 5 Concludes -- References -- Optimization of Stimulus Color for SSVEP-Based Brain-Computer Interfaces in Mixed Reality -- 1 Introduction -- 1.1 A Subsection Sample -- 2 Methods and Materials -- 2.1 Experimental Protocol -- 2.2 Participants -- 2.3 EEG Acquisition and Data Pre-processing -- 2.4 Classification Algorithm -- 2.5 Calculation of Color Contrast -- 3 Result and Analysis -- 4 Conclusion -- References -- Brain Related Research -- White Matter Maturation and Hemispheric Asymmetry During Childhood Based on Chinese Population -- 1 Introduction -- 2 Method -- 2.1 Participants -- 2.2 Image Acquisition -- 2.3 Image Analysis -- 2.4 Tractography Regions of Interest -- 2.5 Statistical Analysis -- 3 Results.
3.1 Age-Related White Matter Maturation -- 3.2 Asymmetry Effects -- 4 Discussion -- 4.1 Age-Related White Matter Maturation -- 4.2 Asymmetry Effects -- 5 Conclusions -- References -- A Digital Gaming Intervention Combing Multitasking and Alternating Attention for ADHD: A Preliminary Study -- 1 Introduction -- 2 Methods -- 2.1 Trial Design -- 2.2 Participants -- 2.3 Randomization and Blinding -- 2.4 Interventions -- 2.5 Outcomes -- 2.6 Data Analysis -- 3 Results -- 3.1 Study Participation -- 3.2 Primary Outcomes -- 3.3 Secondary Outcomes -- 3.4 Adverse Events -- 4 Discussion -- 5 Conclusion -- References -- A BCI Speller with 120 Commands Encoded by Hybrid P300 and SSVEP Features -- 1 Introduction -- 2 Methods and Experiments -- 2.1 Subjects -- 2.2 Hybrid Paradigm Design and Implementation -- 2.3 BCI Experiment -- 2.4 EEG Recording and Processing -- 2.5 Classification Algorithm and Decision Fusion Method -- 2.6 System Performance Evaluation -- 3 Results and Discussion -- 3.1 Subjects EEG Feature Analysis -- 3.2 Offline BCI Performance Analysis -- 3.3 Online BCI Performance Analysis -- 4 Conclusion -- References -- Author Index.