Search Results (12)

This study addresses the selection and application of composite materials for aerospace systems operating in extreme environmental conditions, with a particular focus on high-altitude pseudo-satellites (HAPS). This research is centered on the development of a 400 kg autonomous aerial vehicle (AAV) capable of sustained operations at altitudes of up to 30 km. KMU-3’s microstructure, comprising high-modulus carbon fibers (5–7 µm diameter) in a 5-211B epoxy matrix, provides a high specific strength (1000–2500 MPa), low density (1.6–1.8 g/cm³), and thermal stability (−60 °C to +600 °C), ensuring structural integrity in stratospheric conditions. The mechanical, thermal, and aerodynamic properties of KMU-3-based truss structures were evaluated using finite element method (FEM) simulations, computational fluid dynamics (CFD) analysis, and experimental prototyping. The results indicate that ultra-thin KMU-3 with a wall thickness of 0.1 mm maintains structural integrity under dynamic loads while minimizing overall mass. A novel thermal bonding technique employing 5-211B epoxy resin was developed, resulting in joints with a shear strength of 40 MPa and fatigue life exceeding 10⁶ cycles at 50% load. The material properties remained stable across the operational temperature range of −60 °C to +80 °C. An optimized fiber orientation (0°/90° for longerons and ±45° for diagonals) enhanced the resistance to axial, shear, and torsional stresses, while the epoxy matrix ensures radiation resistance. Finite element method (FEM) and computational fluid dynamics (CFD) analyses, validated by prototyping, confirm the performance of ultra-thin (0.1 mm) truss structures, achieving a lightweight (45 kg) design. These findings provide a validated, lightweight framework for next-generation HAPS, supporting extended mission durations under harsh stratospheric conditions. Full article

Fire detection using deep learning has become a widely adopted approach. However, YOLO-based models often face performance limitations due to NMS, while DETR-based models struggle to meet real-time processing requirements. To address these challenges, we propose FasterGDSF-DETR, a novel fire detection model built upon the RT-DETR framework, designed to enhance both detection accuracy and efficiency. Firstly, this model introduces the FasterDBBNet backbone, which efficiently captures and retains feature information, accelerating the model’s convergence speed. Secondly, we propose the AIFI-GDSF hybrid encoder to reduce information loss in intra-scale interactions and improve the capability of detecting varying morphological flames. Furthermore, to better adapt to complex fire scenarios, we expand the dataset based on the KMU Fire and Smoke database and incorporate WIoU as the loss function to improve model robustness. Experimental results demonstrate that our proposed model surpasses mainstream object detection models in both accuracy and computational efficiency. FasterGDSF-DETR achieves a mean Average Precision of 71.5% on the self-constructed dataset, outperforming the YOLOv9 model of the same scale by 2.4 percentage points. This study introduces a novel task-specific enhancement to the RT-DETR framework, offering valuable insights for future advancements in fire detection technology. Full article

Drivers’ facial expression recognition systems play a pivotal role in Advanced Driver Assistance Systems (ADASs) by monitoring emotional states and detecting fatigue or distractions in real time. However, deploying such systems in resource-constrained environments like vehicles requires lightweight architectures to ensure real-time performance, efficient model updates, and compatibility with embedded hardware. Smaller models significantly reduce communication overhead in distributed training. For autonomous vehicles, lightweight architectures also minimize the data transfer required for over-the-air updates. Moreover, they are crucial for their deployability on hardware with limited on-chip memory. In this work, we propose a novel Dual Attention Lightweight Deep Learning (DALDL) approach for drivers’ facial expression recognition. The proposed approach combines the SqueezeNext architecture with a Dual Attention Convolution (DAC) block. Our DAC block integrates Hybrid Channel Attention (HCA) and Coordinate Space Attention (CSA) to enhance feature extraction efficiency while maintaining minimal parameter overhead. To evaluate the effectiveness of our architecture, we compare it against two baselines: (a) Vanilla SqueezeNet and (b) AlexNet. Compared with SqueezeNet, DALDL improves accuracy by 7.96% and F1-score by 7.95% on the KMU-FED dataset. On the CK+ dataset, it achieves 8.51% higher accuracy and 8.40% higher F1-score. Against AlexNet, DALDL improves accuracy by 4.34% and F1-score by 4.17% on KMU-FED. Lastly, on CK+, it provides a 5.36% boost in accuracy and a 7.24% increase in F1-score. These results demonstrate that DALDL is a promising solution for efficient and accurate emotion recognition in real-world automotive applications. Full article

Background: Ulcerative colitis (UC) is an idiopathic and heterogeneous large intestine disease, characterized by chronic mucosa and submucosa inflammation. Alteration of the intestinal microbiome in UC may be responsible for modifications in metabolite production. Aim: To investigate the microbiota status and trimethylamine-N-oxide (TMAO) metabolite levels in patients with UC according to clinical and endoscopic activity. Methods: As part of a grant project AP14871959 from September 2022 to October 2023, 31 patients with UC and 15 healthy volunteers over 18 years at the Clinic of NCJSC “KMU” were assessed for blood TMAO level and metagenomic sequencing of fecal microbiome. Results: A significant depletion of the main representatives of Bacteroides, Parabacteroides, Prevotella; and an increase in the relative abundance of the genera Actinomyces, Klebsiella, Limosilactobacillus, Streptococcus, Escherichia-Shigella were detected in patients with UC. The number of p_Actinobacteria (g_Collinsella) and p_Eubacterium (g_Xylanophilum) representatives with genes encoding TMA-trimethylamine conversion is significantly reduced in UC patients. TMAO levels were significantly lower in UC patients than in healthy individuals (0.233 µmol/L, p = 0.004). TMAO decreased with disease severity and significantly differed between patients with different activities (p = 0.034). Conclusions: The composition of the intestinal microbiome changes and the level of TMAO decreases in patients with UC at different activities. Full article

The purpose of this study was to investigate the effect that Glycine max hydrolyzed with enzymes from Bacillus velezensis KMU01 has on dextran-sulfate-sodium (DSS)-induced colitis in mice. Hydrolysis improves functional health through the bioconversion of raw materials and increase in intestinal absorption rate and antioxidants. Therefore, G. max was hydrolyzed in this study using a food-derived microorganism, and its anti-inflammatory effect was observed. Enzymatically hydrolyzed G. max (EHG) was orally administered once daily for four weeks before DSS treatment. Colitis was induced in mice through the consumption of 5% (w/v) DSS in drinking water for eight days. The results showed that EHG treatment significantly alleviated DSS-induced body weight loss and decreased the disease activity index and colon length. In addition, EHG markedly reduced tumor necrosis factor-α, interleukin (IL)-1β, and IL-6 production, and increased that of IL-10. EHG improved DSS-induced histological changes and intestinal epithelial barrier integrity in mice. Moreover, we found that the abundance of 15 microorganisms changed significantly; that of Proteobacteria and Escherichia coli, which are upregulated in patients with Crohn’s disease and ulcerative colitis, decreased after EHG treatment. These results suggest that EHG has a protective effect against DSS-induced colitis and is a potential candidate for colitis treatment. Full article

Machine and deep learning techniques are two branches of artificial intelligence that have proven very efficient in solving advanced human problems. The automotive industry is currently using this technology to support drivers with advanced driver assistance systems. These systems can assist various functions for proper driving and estimate drivers’ capability of stable driving behavior and road safety. Many studies have proved that the driver’s emotions are the significant factors that manage the driver’s behavior, leading to severe vehicle collisions. Therefore, continuous monitoring of drivers’ emotions can help predict their behavior to avoid accidents. A novel hybrid network architecture using a deep neural network and support vector machine has been developed to predict between six and seven driver’s emotions in different poses, occlusions, and illumination conditions to achieve this goal. To determine the emotions, a fusion of Gabor and LBP features has been utilized to find the features and been classified using a support vector machine classifier combined with a convolutional neural network. Our proposed model achieved better performance accuracy of 84.41%, 95.05%, 98.57%, and 98.64% for FER 2013, CK+, KDEF, and KMU-FED datasets, respectively. Full article

Monitoring drivers’ emotions is the key aspect of designing advanced driver assistance systems (ADAS) in intelligent vehicles. To ensure safety and track the possibility of vehicles’ road accidents, emotional monitoring will play a key role in justifying the mental status of the driver while driving the vehicle. However, the pose variations, illumination conditions, and occlusions are the factors that affect the detection of driver emotions from proper monitoring. To overcome these challenges, two novel approaches using machine learning methods and deep neural networks are proposed to monitor various drivers’ expressions in different pose variations, illuminations, and occlusions. We obtained the remarkable accuracy of 93.41%, 83.68%, 98.47%, and 98.18% for CK+, FER 2013, KDEF, and KMU-FED datasets, respectively, for the first approach and improved accuracy of 96.15%, 84.58%, 99.18%, and 99.09% for CK+, FER 2013, KDEF, and KMU-FED datasets respectively in the second approach, compared to the existing state-of-the-art methods. Full article

Bacillus velezensis strain KMU01 showing γ-glutamyltransferase activity as a probiotic candidate was isolated from kimchi. However, the genetic information on strain KMU01 was not clear. Therefore, the current investigation was undertaken to prove the probiotic traits of B. velezensis strain KMU01 through genomic analysis. Genomic analysis revealed that strain KMU01 did not encode enterotoxin genes and acquired antibiotic resistance genes. Strain KMU01 genome possessed survivability traits under extreme conditions such as in the presence of gastric acid, as well as several probiotic traits such as intestinal epithelium adhesion and the production of thiamine and essential amino acids. Potential genes for human health enhancement such as those for γ-glutamyltransferase, nattokinase, and bacteriocin production were also identified in the genome. As a starter candidate for food fermentation, the genome of KMU01 encoded for protease, amylase, and lipase genes. The complete genomic sequence of KMU01 will contribute to our understanding of the genetic basis of probiotic properties and allow for the assessment of the effectiveness of this strain as a starter or probiotic for use in the food industry. Full article

Protein kinases regulate protein phosphorylation, which are involved in fundamental cellular processes such as inflammatory response. In this study, we discovered a novel multi-protein kinase inhibitor, KMU-1170, a derivative of indolin-2-one, and investigated the mechanisms of its inflammation-inhibiting signaling in both THP-1 cells and human osteoarthritic fibroblast-like synoviocytes (FLS). We demonstrated that in THP-1 cells, KMU-1170 inhibited lipopolysaccharide (LPS)-induced upregulation of inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2), and, furthermore, suppressed LPS-induced phosphorylation of transforming growth factor-β-activated kinase 1, JNK, ERK, inhibitor of NF-κB kinase α/β (IKKα/β), and NF-κB p65 as well as nuclear translocation of NF-κB p65. Moreover, KMU-1170 suppressed LPS-induced upregulation of proinflammatory cytokines such as IL-1β, TNF-α, and IL-6, and, notably, inhibited LPS-induced upregulation of the NLRP3 inflammasome in THP-1 cells. Importantly, KMU-1170 attenuated LPS-mediated inflammatory responses in human osteoarthritic FLS, such as the upregulation of IL-1β, TNF-α, IL-6, iNOS, and COX-2 and the phosphorylation of IKKα/β and NF-κB p65. Collectively, these results suggest that KMU-1170 inhibits inflammatory signal transduction and could be developed as a potential anti-inflammatory agent. Full article

PIM kinases, a small family of serine/threonine kinases, are important intermediates in the cytokine signaling pathway of inflammatory disease. In this study, we investigated whether the novel PIM kinase inhibitor KMU-470, a derivative of indolin-2-one, inhibits lipopolysaccharide (LPS)-induced inflammatory responses in RAW 264.7 cells. We demonstrated that KMU-470 suppressed the production of nitric oxide and inducible nitric oxide synthases that are induced by LPS in RAW 264.7 cells. Furthermore, KMU-470 inhibited LPS-induced up-regulation of TLR4 and MyD88, as well as the phosphorylation of IκB kinase and NF-κB in RAW 264.7 cells. Additionally, KMU-470 suppressed LPS-induced up-regulation at the transcriptional level of various pro-inflammatory cytokines such as IL-1β, TNF-α, and IL-6. Notably, KMU-470 inhibited LPS-induced up-regulation of a major component of the inflammasome complex, NLRP3, in RAW 264.7 cells. Importantly, PIM-1 siRNA transfection attenuated up-regulation of NLRP3 and pro-IL-1β in LPS-treated RAW 264.7 cells. Taken together, these findings indicate that PIM-1 plays a key role in inflammatory signaling and that KMU-470 is a potential anti-inflammatory agent. Full article

Pedestrian-related accidents are much more likely to occur during nighttime when visible (VI) cameras are much less effective. Unlike VI cameras, infrared (IR) cameras can work in total darkness. However, IR images have several drawbacks, such as low-resolution, noise, and thermal energy characteristics that can differ depending on the weather. To overcome these drawbacks, we propose an IR camera system to identify pedestrians at night that uses a novel attention-guided encoder-decoder convolutional neural network (AED-CNN). In AED-CNN, encoder-decoder modules are introduced to generate multi-scale features, in which new skip connection blocks are incorporated into the decoder to combine the feature maps from the encoder and decoder module. This new architecture increases context information which is helpful for extracting discriminative features from low-resolution and noisy IR images. Furthermore, we propose an attention module to re-weight the multi-scale features generated by the encoder-decoder module. The attention mechanism effectively highlights pedestrians while eliminating background interference, which helps to detect pedestrians under various weather conditions. Empirical experiments on two challenging datasets fully demonstrate that our method shows superior performance. Our approach significantly improves the precision of the state-of-the-art method by 5.1% and 23.78% on the Keimyung University (KMU) and Computer Vision Center (CVC)-09 pedestrian dataset, respectively. Full article

In recent years, researchers of deep neural networks (DNNs)-based facial expression recognition (FER) have reported results showing that these approaches overcome the limitations of conventional machine learning-based FER approaches. However, as DNN-based FER approaches require an excessive amount of memory and incur high processing costs, their application in various fields is very limited and depends on the hardware specifications. In this paper, we propose a fast FER algorithm for monitoring a driver’s emotions that is capable of operating in low specification devices installed in vehicles. For this purpose, a hierarchical weighted random forest (WRF) classifier that is trained based on the similarity of sample data, in order to improve its accuracy, is employed. In the first step, facial landmarks are detected from input images and geometric features are extracted, considering the spatial position between landmarks. These feature vectors are then implemented in the proposed hierarchical WRF classifier to classify facial expressions. Our method was evaluated experimentally using three databases, extended Cohn-Kanade database (CK+), MMI and the Keimyung University Facial Expression of Drivers (KMU-FED) database, and its performance was compared with that of state-of-the-art methods. The results show that our proposed method yields a performance similar to that of deep learning FER methods as 92.6% for CK+ and 76.7% for MMI, with a significantly reduced processing cost approximately 3731 times less than that of the DNN method. These results confirm that the proposed method is optimized for real-time embedded applications having limited computing resources. Full article