Daegu Gyeongbuk Institute of Science and Technology
DGIST Library Institutional RepositoryNot a member yet
12664 research outputs found
Sort by
인간의 시간 추정과 관련된 신경 기저
No full texttime estimation, contextual information, fMRI, Bayesian modeling, functional connectivity, dynamic time warping시간 추정은 동적 환경에서 의사결정, 행동 조정 및 적응적 행동을 가능하게 하는 중요한 인지 기능이다. 본 논문은 시간 추정의 인지적 및 신경적 메커니즘을 심층적으로 탐구하며, 특히 맥락적 정보의 역할에 초점을 맞춘다. 행동 실험과 fMRI 연구를 통해 사전 맥락 정보를 제공받은 참가자들이 시간 추정의 정확도와 반응 전략에 미치는 영향을 분석하고, 이러한 효과가 뇌의 신경 활동에 어떻게 반영되는지를 조사하였다.
행동 실험 결과, 사전 맥락 정보를 제공받은 참가자들은 시간 추정의 정확도가 더 높았으며, 맥락에 따라 반응을 효과적으로 조정하는 모습을 보였다. fMRI 분석에서는 맥락 정보를 처리하는 뇌 영역이 예상적 역할을 수행하며, 시간 처리와 관련된 영역보다 먼저 활성화되는 것을 보여주었다. 기능적 연결성 분석에서는 사전 맥락 정보를 제공받은 그룹과 그렇지 않은 그룹 간의 신경 적응 패턴 차이가 드러나, 맥락적 정보가 뇌 네트워크 동역학을 조절함을 시사했다. Dynamic Time Warping 분석 결과, 사전 맥락 정보를 제공받은 그룹은 관련 뇌 영역에서 더 긴 신경 활동 지속 시간을 유지하여 예상적 처리와 맥락 단서 통합의 효율성을 높이고, 시간 추정의 정확도를 향상시키는 것으로 나타났다.
본 연구는 맥락적 정보가 시간 추정 과정에서 행동 및 신경 반응을 조절하는 메커니즘을 밝혀내어, 시간 지각에 관한 인지 모델을 한층 풍부하게 이해하는 데 중요한 통찰을 제공합니다.|Time estimation is a vital cognitive function that underpins decision-making, action coordination, and adaptive behavior in dynamic environments. This thesis investigates the cognitive and neural mechanisms underlying time estimation, with a particular emphasis on the role of contextual information. Through a series of behavioral experiments and fMRI studies, we examine how contextual awareness influences time estimation accuracy and response strategies, and how these effects manifest in neural activity.
Behavioral findings indicate that participants with prior contextual knowledge produce more accurate time estimates and adapt their responses effectively according to the context. FMRI analyses reveal that brain regions involved in processing contextual information activate in advance of regions associated with processing duration, highlighting an anticipatory role of contextual information-related regions in time estimation. Functional connectivity analyses further demonstrate distinct neural adaptation patterns between the Informed group, who had access to contextual information, and the Uninformed groups, who did not, illustrating how contextual information shapes brain network dynamics. The dynamic time warping results show that the Informed group sustains prolonged neural activity in regions associated with contextual information, indicating enhanced anticipatory processing and more efficient integration of contextual cues.
These findings provide novel insights into how contextual information modulates both behavioral and neural responses in time estimation. This research advances the understanding of how the brain integrates contextual cues with temporal processing, enriching cognitive models of time perception.Ⅰ. Introduction 1
1.1 Preface 1
1.2 Research Background 2
1.2.1 Cognitive Models of Time Perception 2
1.2.1.1 Scalar Expectancy Theory 2
1.2.1.2 Bayesian Models in Time Estimation: The Role of Prior Knowledge 6
1.2.1.3 Iterative Bayesian Model: Dynamic Prior Updating 6
1.2.2 Neural Underpinnings of Duration and Contextual Information Processing 8
1.2.2.1 Duration Processing 8
1.2.2.2 Contextual Information Processing 9
1.2.2.3 Implicit vs. Explicit Contextual Influence 11
1.3 Limitations of Previous Research 12
1.4 Hypothesis 14
1.5 Overview of the Thesis 15
ⅠⅠ. Experiment 1 - Behavioral study 17
2.1 Methods 17
2.1.1 Participants 17
2.1.2 Procedure 17
2.1.2.1 Task Design 17
2.1.2.2 Psychological Assessments 21
2.1.2.3 Post Survey 21
2.1.3 Data Analysis 26
2.1.3.1 Data Preprocessing 26
2.1.3.2 Bayesian Linear Regression Model for Estimation Error 27
2.1.3.3 Kalman Filter-based Iterative Bayesian Model for Time Estimation 29
2.2 Results 34
2.2.1 Bayesian Linear Regression Model Analysis: Contextual Awareness Enhances Accuracy in Time
Estimation 34
2.2.2 Kalman Filter-based Iterative Bayesian Model: Contextual Information Drives Response Bias
Adjustments in Time Estimation 40
2.3 Discussion 44
ⅠⅠⅠ. Experiment 2 - fMRI Study 47
3.1 Methods 47
3.1.1 Participants 47
3.1.2 Procedure 48
3.1.3 MRI acquisition 49
3.1.4 Data analysis 50
3.1.4.1 Data preprocessing 50
3.1.4.2 Bayesian Linear Regression Model for fMRI Study 52
3.1.4.3 Kalman Filter-Based Iterative Bayesian Model for Time Estimation in fMRI Study 53
3.1.4.4 Seed-based functional connectivity 55
3.1.4.5 Dynamic Time Warping Analysis of Contextual Information's Impact on Time Estimation 57
3.2 Results 60
3.2.1 Contextual Information and Its Role in Modulating Time Estimation Biases 60
3.2.2 Neural Dynamics in Time Estimation: Multi-Factor Influences in fMRI Analysis 64
3.2.3 Impact of Contextual Information on Response Shifts 70
3.2.4 Influence of Contextual Information on Time Estimation Through BOLD Signal in Iterative Bayesian
Model 72
3.2.5 Functional Connectivity Analysis: Right Caudate Nucleus’s Role in Contextual Integration in Time
Estimation 76
3.2.6 Dynamic Time Warping Analysis Reveals Prolonged Neural Engagement in Contextual Processing for
Informed Group 79
3.3 Discussion 84
IV. Conclusion 90
References 94
요약문 100DoctordCollectio
라스트 마일 배달 서비스를 위한 착륙 유형 기반 다중 드론 경로 생성
No full textDrone delivery, Route Planning, Landing Type, Mixed Integer Linear Programming (MILP), Heuristic Algorithm1. Introduction 1
2. Related Work 4
2.1 Drone Route Planning 4
2.2 Commercial Drones’ Landing Behavior 4
3. Problem Statement 6
4. Approach Overview 8
5. Optimal Multi-Drone Route Generation with Landing Exclusion Zone 10
5.1 Constraints 10
5.1.1 Constraint 1: Drone Dynamics 10
5.1.2 Constraint 2: Collision Avoidance 11
5.1.3 Constraint 3: Delivery Requirements 11
5.1.4 Constraint 4: Landing Exclusion Zone 12
5.2 Objective 13
6. Landing-Aware Heuristic Algorithm 15
6.1 Drone Order Sorting 15
6.1.1 Hypothesis 1 15
6.1.2 Hypothesis 2 16
6.2 Landing-Aware Heuristic Algorithm 17
6.3 Property of Heuristic Algorithm 19
7. Experiment and Case Study 20
7.1 Experiment Setting 20
7.2 Results 20
7.2.1 Delivery Time and Computing Time 20
7.2.2 Effect of Landing Exclusion Zones 21
7.2.3 Effect of Drone Ordering 21
7.2.4 Case Study 23
8. Conclusion 24
9. Appendix 25
9.1 Proof of Lemma 1 25
9.2 Proof of Lemma 2 27
9.3 Proof of Lemma 3 29
9.4 Proof of Lemma 4 29
References 31MasterdCollectio
RotOptiLink: A Novel Mechanism for Minimizing Blind Spots with an Integrated Rotational Optical Sensing Device in Manipulator Links
No full textAs robotics technology advances and the fifth industrial revolution unfolds, bringing robots closer to human life, research on ensuring human safety in shared workspaces has become crucial. Among these studies, manipulators, which are widely used in various fields, have gained prominence. Existing research predominantly employs sensor ring and sensor skin types for safety assurance by measuring robot-human distances. Sensor rings, with fewer sensors and lower costs, suffer from blind spots. Conversely, sensor skin types, despite being blind spot-free, incur high computational and fabrication expenses. Our proposed solution, RotOptiLink, integrates a rotational optical sensing device into manipulator links, combining the benefits of both types while mitigating their drawbacks. RotOptiLink utilizes a slip ring mechanism, enabling the outer ring, equipped with distance sensors, to rotate while maintaining contact with the inner ring, which carries power and signal processing lines. This setup allows for the detection of approaching individuals and obstacles without blind spots, even with a small number of sensors. We mathematically generalize design parameters, validate them experimentally, and demonstrate omnidirectional obstacle sensing on a three-axis manipulator. © IEEE.FALSEsciescopu
새로이 설계된 형광 단백질을 활용한 리간드-수용체 상호작용의 정량화
No full textBiFC, Erythropoietin receptor, Protein engineering, Protein modellingI. Introduction 1
II. Materials and methods 4
1. Designing biosensor system with spGFP assembly system 4
2. Transformation, culture and overexpression on e.coli system 4
3. Purification of target protein 5
4. Normalizing molar ration of sensor molecules 6
5. Detection of signal emission from biosensor system. 7
6. Cloning of target gene 7
7. Design of novel peptide ligand of erythropoietin receptor 8
8. Statistical analysis 9
III. Results 10
1. Biosensor property configuration 10
2. Biosensor activity confirmation 12
3. Evaluating activity of denatured EPO 15
4. Testing EPO’s interaction between spGFP domains 17
5. Cloning another biosensor with different receptor system 19
6. Ligand design with RFdiffusion – single binding peptides 21
7. Dimerization screening of designed ligands by BiFC – single binding peptides 23
8. Ligand design with RFdiffusion – dual binding peptides 25
9. Dimerization screening of designed ligands by BiFC – dual binding peptides 26
IV. Discussion 30
V. Conclusion 33
VI. References 35MasterdCollectio
마그네슘 배터리 양극 물질로서 무수 프러시안 블루 유사체의 합성과 전기화학적 특성 평가
No full textPrussian blue analogues, Cathode materials, Multivalent-ion batteries, Magnesium batteriesⅠ. Introduction
1.1. Why rechargeable magnesium batteries (RMBs)? 1
1.2. Challenges in cathode development for RMBs 2
1.3. Prussian Blue analogues (PBAs) as host materials 3
Ⅱ. Experimental section
2.1. Material synthesis 5
2.2. Material Characterization 6
2.3. Electrochemical Characterization 6
2.4. Galvanostatic intermittent titration technique (GITT) analysis 8
2.5. Structural analysis 8
Ⅲ. Results and Discussion
3.1. Synthesis of iron hexacyanoferrate 9
3.2. Electrochemical characterization 18
3.3. Mg2+ intercalation mechanism 22
3.4. Compatibility of FeHCF with a Mg metal anode 26
IV. Conclusion 28
V. References 29
요약문MasterdCollectio
GAN 기반 합성 초음파 이미지 생성: 데이터 부족 및 도메인 적응 문제 해결
No full textDeep Learning, Data Augmentation, Domain Adaptation, Ultrasound Image, CycleGANⅠ. Introduction 1
1.1 Background 1
1.2 Related Work 2
1.3 Research Objectives and Approach 7
II. Methods 10
2.1 Network 10
2.1.1 Model Overview 10
2.1.2 Framework 11
2.1.3 Formulation 12
2.2 Datasets 15
2.2.1 Open-Source Datasets 16
2.2.2 Field II Simulation 19
2.3 Training Setup and Details 21
2.3.1 Network Architecture 21
2.3.2 Loss Function 25
2.3.3 Training Setup 27
III. Results 29
3.1 CycleGAN Results 29
3.2 Evaluation 34
3.2.1 Evaluation Metrics 35
3.2.2 Evaluation Method 37
3.2.3 Training Setup 38
3.2.4 Evaluation Results 40
IV. Conclusion 48
4.1 Future Works 48
References 49
국문요약 53MasterdCollectio
Bi-layer Structured Metal Fluoride/Polymer Protection of Zn-metal Anode for Aqueous Zn-ion Batteries
No full textOne-step Zn modification, Bi-layer structured protection layer, Uniform Zn deposition, HER suppression, Aqeuous Zn-ion batteriesAbstract i
List of contents ii
List of table iii
List of figures iv
Ⅰ. Introduction 1
1.1. Historical Background of Aqueous Zn metal batteries (AZBs) 1
1.2. Principles of AZBs 2
1.3. Limitations of AZBs and previous works 3
1.4. Research Goals 4
II. Results and discussion 5
2.1. Design and preparation of BSPL@Zn electrode 5
2.2. Synergistic effect of BSPL@Zn for Zn deposition 6
2.3. Free-water access inhibition of BSPL@Zn 7
2.4. Investigating BSPL@Zn for electrochemical stability 8
2.5. NH4V4O10 (NVO) cathode based full cell test 10
III. Conclusion 12
IV. Experimental Method 13
4.1. Preparation of protected ZMAs (AgF@Zn and BSPL@Zn) 13
4.2. Synthesis of NVO and cathode preparation 13
4.3. Electrochemical test 14
4.4. Characterization 14
References 49
Summary in Korean 54MasterdCollectio
STSPhys: Enhanced Remote Heart Rate Measurement With Spatial-Temporal SwiftFormer
No full textEstimating heart activities and physiological signals from facial video without any contact, known as remote photoplethysmography and remote heart rate estimation, holds significant potential for numerous applications. In this letter, we present a novel approach for remote heart rate measurement leveraging a Spatial-Temporal SwiftFormer architecture (STSPhys). Our model addresses the limitations of existing methods that rely heavily on 3D CNNs or 3D visual transformers, which often suffer from increased parameters and potential instability during training. By integrating both spatial and temporal information from facial video data, STSPhys achieves robust and accurate heart rate estimation. Additionally, we introduce a hybrid loss function that integrates constraints from both the time and frequency domains, further enhancing the model's accuracy. Experimental results demonstrate that STSPhys significantly outperforms existing state-of-the-art methods on intra-dataset and cross-dataset tests, achieving superior performance with fewer parameters and lower computational complexity. © IEEE.FALSEsci
BEE-SLAM: A 65-nm 17.96-TOPS/W Location-Sharing-Based Multi-Agent Neuromorphic SLAM Accelerator for Swarm Robotics
No full textMulti-agent (MA) simultaneous localization and mapping (SLAM) has been rigorously explored to enhance map accuracy in swarm robotics. Although centralized MA SLAM systems, which depend on a server for complex computations in map optimization, have been extensively studied, the circuit-domain approaches to decentralized MA SLAM systems are still limited due to challenges such as limited memory capacity and security vulnerabilities in wireless inter-agent data transmission. Thus, we propose a BEE-SLAM accelerator, a location-sharing MA neuromorphic SLAM accelerator inspired by bee communication for decentralized MA SLAM systems. The location-sharing-based MA error correction (MAEC) is employed to attain accurate map results without loop closure with a 94.81% reduced number of operations compared to the global map-based MA SLAM. In addition, a 7 × 7 pulsewidth modulation (PWM)-based hybrid mixed-signal/digital pose-cell (HY-PC) array with pseudo pose cells (PPCs) achieves 2.04 × energy efficiency compared to the oscillatory pose-cell array. The test chip fabricated in a 65-nm CMOS technology achieves a peak energy efficiency of 17.96 TOPS/W under 350 × 450 m outdoor exploration. © IEEE.FALSEsciescopu