Kyutacar : Kyushu Institute of Technology Academic Repository
Not a member yet
9746 research outputs found
Sort by
Large-scale moral machine experiment on large language models
Full text linkThe rapid advancement of Large Language Models (LLMs) and their potential integration into autonomous driving systems necessitates understanding their moral decision-making capabilities. While our previous study examined four prominent LLMs using the Moral Machine experimental framework, the dynamic landscape of LLM development demands a more comprehensive analysis. Here, we evaluate moral judgments across 52 different LLMs, including multiple versions of proprietary models (GPT, Claude, Gemini) and open-source alternatives (Llama, Gemma), to assess their alignment with human moral preferences in autonomous driving scenarios. Using a conjoint analysis framework, we evaluated how closely LLM responses aligned with human preferences in ethical dilemmas and examined the effects of model size, updates, and architecture. Results showed that proprietary models and open-source models exceeding 10 billion parameters demonstrated relatively close alignment with human judgments, with a significant negative correlation between model size and distance from human judgments in open-source models. However, model updates did not consistently improve alignment with human preferences, and many LLMs showed excessive emphasis on specific ethical principles. These findings suggest that while increasing model size may naturally lead to more human-like moral judgments, practical implementation in autonomous driving systems requires careful consideration of the trade-off between judgment quality and computational efficiency. Our comprehensive analysis provides crucial insights for the ethical design of autonomous systems and highlights the importance of considering cultural contexts in AI moral decision-making.journal articl
電力線通信技術を用いた革新的な部分放電監視システムの開発
Full text link九州工業大学九州工業大学博士学位論文（要旨）学位記番号: 工博甲第602号 学位授与年月日:令和7年3月25日thesi
High-entropy engineering enables O3-type layered oxide with high structural stability and reaction kinetic for sodium storage
No full textO3-type layered oxides are considered promising cathode materials for sodium-ion batteries (SIBs) due to their high theoretical capacity, but they often face issues with structural instability and poor sodium-ion diffusion, leading to rapid capacity fading. In this work, we introduce a high-entropy approach combined with synergistic multi-metal effects to address these limitations by enhancing both the structural stability and reaction kinetics. A novel O3-type layered high-entropy cathode material, Na0.9Fe0.258Co0.129Ni0.258Mn0.258Ti0.097O2 (TMO5), which was synthesized via a straightforward solid-phase method for easy mass production. Experimental analysis combined with in/ex-situ characterization verifies that high-entropy metal ion mixing contributes to the improved reversibility of the redox reaction and O3-P3-O3 phase transition behaviors, as well as the enhanced Na+ diffusivity. Benefit from the advantage of structure and composition, the TMO5 exhibits a higher initial specific capacity of 159.6 mAh g−1 and an impressive capacity retention of 85.6 % after 100 cycles at 2 C with the specific capacity of 110.1 mAh g−1. This work showcases high-entropy O3-type layered oxides as a promising pathway for achieving robust, high-performance SIB cathodes.journal articl
The Digital Transformation in Sustainable Agriculture
Full text link九州工業大学博士（情報工学）1 Introduction| 2 AgricultureinDigitalEra| 3 FuzzyLogicinEvolvingTraditionalFarmingtoSmartAgriculture| 4 FromSoiltoCloud:FutureDirectionsonAgri-DigitalTransformation| 5 BlockchainTowardSmartAgricultureRevolutions| 6 Data-DrivenSmartFarming:IntegratingEdgeComputing,Blockchain,and Gen-AI| 7 ConclusionandFutureWork106 Bibliography九州工業大学博士学位論文 学位記番号:情工博甲第411号 学位授与年月日:令和7年6月30日令和7年度doctoral thesi
Spatially Nonuniform Filtering and Encoding of Visual Information in a Retinomorphic Hardware System
Full text linkEfficient information processing in biological early-vision systems is thought to rely on spatially nonuniform computations in the central and peripheral visual fields. Inspired by this functional segregation, we developed a digital hardware vision system that integrates two distinct spatiotemporal filters and spike encoders for the central and peripheral regions on a single image sensor plane. The system employs digital processing techniques to emulate graded potential signals and point-process spike outputs, mimicking the functionality of vertebrate retinas. System-level testing on actual hardware equipment verified region-specific response characteristics for the central and peripheral regions. This proposed approach advances the realization of retinomorphic efficient computations in future robotics and embedded vision systems.journal articl
Preparation of activated carbon derived from bamboo treated with hot-compressed water or saturated steam and performance as electrodes of supercapacitor
Full text link九州工業大学博士（工学）1 Introduction and Context| 2 Supercapacitor electrode made from hot-compressed water-treated bamboo| 3 Supercapacitor electrode made from saturated steam-treated bamboo| 4 Activated carbon by alkaline compounds as supercapacitor electrode| 5 Activated carbon by alkali metal hydroxide as supercapacitor electrode| 6 General Conclusions and Prospects九州工業大学博士学位論文 学位記番号:工博甲第613号 学位授与年月日:令和7年9月25日令和7年度doctoral thesi
Conversion of Low-Value Biomass into High-Value Carbon Materials for Energy Storage and Environmental Applications
Full text link九州工業大学博士（工学）1 GENERAL INTRODUCTION AND REVIEW OF LITERATURE| 2 BIOCHAR AND H₃PO₄-ACTIVATED CARBON DERIVED FROM DICRANOPTERIS LINEARIS FERN: ELECTROCHEMICAL BEHAVIOR IN ELECTRIC DOUBLE-LAYER CAPACITOR APPLICATIONS| 3 ELECTRIC DOUBLE-LAYER CAPACITORS UTILIZING ACTIVATED CARBONS DERIVED FROM JAPANESE CEDAR BARK WITH DIFFERENT CHEMICAL ACTIVATORS| 4 UTILIZATION OF BAMBOO AND CEDAR BARK RESIDUES FOR BIOCHAR PRODUCTION AS CO₂ AND RADIONUCLIDE ADSORBENTS| 5 KOH ACTIVATION REVIEW: CHEMICAL REACTIONS| 6 CONCLUSIONS AND OUTLOOK九州工業大学博士学位論文 学位記番号:工博甲第614号 学位授与年月日:令和7年9月25日令和7年度doctoral thesi
Allowable Bending Stress for Glass Beams: Experiments and Statistical Analysis
Full text linkA fundamental understanding of resisting and collapse mechanisms is required to ensure the structural safety of primary load-bearing elements, like columns and beams. The present research, in this regard, experimentally investigates the in-plane bending and Lateral-Torsional Buckling (LTB) response of tempered glass beams, analyzing key factors such as the evolution of bending stress, the critical buckling moment, and the corresponding failure mode for a total of 52 original samples. The study focuses on specimens with small slenderness (less than 1.2), revealing that well-defined failure mechanisms can be detected when varying rhis ratio. The stress ratio (r = 0.57) is crucial in determining the bending behaviour, with high slenderness following Euler’s buckling curve for beams in LTB, but medium- or low-sleenderness specimens exhibiting an inelastic response, with a sudden and brittle collapse mechanism which diverges from Euler’s curve, indicating a critical condition for safety. The study identifies three slenderness ratios satisfying specific limits. Statistical analysis is applied to the collected results, revealing that the allowable stress curve begins on the y-axis with an approximate reduction factor of 0.55 and converges with Euler’s curve at high slenderness values. As shown, the strategy can be used to correlate the failure type to the slenderness. It could hence provide practical support for improving the safety and efficiency levels of structural glass members. Besides, a more extensive database will be required for a sound and robust calibration of key parameters.journal articl
Detecting Phagocytotic Activity of Leukocytes in Gram Stained Smears Images
No full textIn this paper, we detect phagocytotic activity of leukocytes from Gram stained smear images by using object detectors YOLOv5 and YOLOv8. Then, we detect three kinds of regions representing phagocytotic, quasi-phagocytotic and non-phagocytotic activities and two kinds of regions representing phagocytotic and non-phagocytotic containing quasi-phagocytotic activities. Furthermore, we apply the several methods of data augmentation to the detection and investigate the effect of them.journal articl
Associations of Vocal Features, Psychiatric Symptoms, and Cognitive Functions in Schizophrenia
Full text linkPurpose: This study explored the use of advanced computational techniques in vocal analysis to improve the assessment of psychiatric symptoms and cognitive functions in schizophrenia. We hypothesized that digital signal processing techniques, such as mel spectrogram and mel-frequency cepstral coefficients (MFCC), could be used for objective evaluation of psychiatric symptoms and cognitive functions based on the analysis of alterations in the vocal characteristics.
Patients and Methods: Voice samples from 14 participants diagnosed with schizophrenia (92.9% female) were collected using a microphone array, and vocal features were extracted from the samples using mel spectrogram and MFCC techniques. Psychiatric symptoms and cognitive functions were assessed using the Positive and Negative Syndrome Scale (PANSS) and the computer-based tool Cognitrax.
Results: We found significant negative correlations between specific vocal features (mel spectrogram and MFCC) and cognitive functions, particularly working memory (β = − 0.645, p = 0.023) and sustained attention (β = − 0.626, p = 0.029). No direct correlations were found between vocal features and psychiatric symptoms, as measured by PANSS scores. However, the correlations between cognitive functions and PANSS total scores were significant (β = − 0.604, p = 0.037), suggesting that cognitive functions may mediate the relationship between psychiatric symptoms and vocal characteristics.
Conclusion: This study underscores the potential of vocal analysis as a non-invasive tool for assessing cognitive impairment in schizophrenia. Future research should focus on expanding the sample size and including diverse populations to enhance the generalizability of these findings.
Plain Language Summary: Why was the study done? Schizophrenia is a complex mental health disorder that affects how people think, feel, and behave, often leading to challenges in communication and cognitive function. This study aimed to explore how changes in vocal features can relate to the severity of psychiatric symptoms and cognitive abilities in individuals diagnosed with schizophrenia. We wanted to fill a gap in existing research about the connections between vocal patterns and cognitive functions in people with schizophrenia.
What did the researchers do and find? We analyzed voice data from 14 participants to see if there were links between their vocal features, psychiatric symptoms, and cognitive functions. Our findings showed that poorer cognitive abilities, like memory and attentional control, negatively correlated with certain vocal characteristics. Additionally, higher psychiatric symptom severity was associated with worse cognitive performance.
What do these results mean? This study suggests that analyzing vocal features could help us better understand cognitive impairments in schizophrenia, offering new ways to assess patients. By looking at how vocal features and cognitive functions relate, we can develop improved evaluation methods that may lead to better treatment outcomes. This research adds to our understanding of schizophrenia and highlights the importance of using vocal analysis in future studies to enhance clinical assessments and interventions.journal articl