Woongkyu Park¹, Yeongyu Choi², Mahammad Shareef Mekala³, Gyu Sang Choi¹, Kook-Yeol Yoo¹, Ho-youl Jung^1,*

CMC-Computers, Materials & Continua, Vol.82, No.3, pp. 3965-3981, 2025, DOI:10.32604/cmc.2025.059966 - 06 March 2025

Abstract Designing fast and accurate neural networks is becoming essential in various vision tasks. Recently, the use of attention mechanisms has increased, aimed at enhancing the vision task performance by selectively focusing on relevant parts of the input. In this paper, we concentrate on squeeze-and-excitation (SE)-based channel attention, considering the trade-off between latency and accuracy. We propose a variation of the SE module, called squeeze-and-excitation with layer normalization (SELN), in which layer normalization (LN) replaces the sigmoid activation function. This approach reduces the vanishing gradient problem while enhancing feature diversity and discriminability of channel attention. In… More >

Can Hu¹, Shanqing Zhang^2,*, Kewei Tao², Gaoming Yang¹, Li Li²

CMC-Computers, Materials & Continua, Vol.82, No.3, pp. 4913-4930, 2025, DOI:10.32604/cmc.2025.059770 - 06 March 2025

Abstract The surge of large-scale models in recent years has led to breakthroughs in numerous fields, but it has also introduced higher computational costs and more complex network architectures. These increasingly large and intricate networks pose challenges for deployment and execution while also exacerbating the issue of network over-parameterization. To address this issue, various network compression techniques have been developed, such as network pruning. A typical pruning algorithm follows a three-step pipeline involving training, pruning, and retraining. Existing methods often directly set the pruned filters to zero during retraining, significantly reducing the parameter space. However, this… More >

Abubakar Unguwanrimi Yakubu^1,2,4, Jiahao Zhao¹, Qi Jiang¹, Xuanhong Ye¹, Junyi Liu¹, Qinglong Yu¹, Shusheng Xiong^1,3,4,*

Energy Engineering, Vol.122, No.3, pp. 1025-1051, 2025, DOI:10.32604/ee.2025.057680 - 07 March 2025

Abstract This study uses numerical simulations of liquid cooling flow fields to investigate polymer exchange membrane fuel cell (PEMFC) thermal control. The research shows that the optimum cooling channel design significantly reduces the fuel cell’s temperature differential, improving overall efficiency. Specifically, the simulations show a reduction in the maximum temperature by up to 15% compared to traditional designs. Additionally, according to analysis, the Nusselt number rises by 20% with the implementation of serpentine flow patterns, leading to enhanced heat transfer rates. The findings demonstrate that effective cooling strategies can lead to a 10% increase in fuel More >

Abdullah M. Alqahtani^1,*, Kamran Ahmad Awan², Abdulaziz Almaleh³, Osama Aletri⁴

CMES-Computer Modeling in Engineering & Sciences, Vol.142, No.3, pp. 3155-3179, 2025, DOI:10.32604/cmes.2025.061472 - 03 March 2025

Abstract Efficient resource management within Internet of Things (IoT) environments remains a pressing challenge due to the increasing number of devices and their diverse functionalities. This study introduces a neural network-based model that uses Long-Short-Term Memory (LSTM) to optimize resource allocation under dynamically changing conditions. Designed to monitor the workload on individual IoT nodes, the model incorporates long-term data dependencies, enabling adaptive resource distribution in real time. The training process utilizes Min-Max normalization and grid search for hyperparameter tuning, ensuring high resource utilization and consistent performance. The simulation results demonstrate the effectiveness of the proposed method, More >

Gaoshuai Su^1,2, Hongyang Li^1,*, Huafeng Chen^1,2

CMES-Computer Modeling in Engineering & Sciences, Vol.142, No.3, pp. 2527-2545, 2025, DOI:10.32604/cmes.2025.059452 - 03 March 2025

Abstract Brain tumors, one of the most lethal diseases with low survival rates, require early detection and accurate diagnosis to enable effective treatment planning. While deep learning architectures, particularly Convolutional Neural Networks (CNNs), have shown significant performance improvements over traditional methods, they struggle to capture the subtle pathological variations between different brain tumor types. Recent attention-based models have attempted to address this by focusing on global features, but they come with high computational costs. To address these challenges, this paper introduces a novel parallel architecture, ParMamba, which uniquely integrates Convolutional Attention Patch Embedding (CAPE) and the… More >

Abdelilah Makaoui¹, El Bachir Lahmer^1,*, Jaouad Benhamou^1,2, Mohammed Amine Moussaoui¹, Ahmed Mezrhab¹

Frontiers in Heat and Mass Transfer, Vol.23, No.1, pp. 207-230, 2025, DOI:10.32604/fhmt.2024.060166 - 26 February 2025

Abstract This study focuses on numerically investigating thermal behavior within a differentially heated cavity filled with nanofluid with and without obstacles. Numerical comparison with previous studies proves the consistency and efficacy of the lattice Boltzmann method associated with a single relaxation time and its possibility of studying the nanofluid and heat transfer with high accuracy. Key parameters, including nanoparticle type and concentration, Rayleigh number, fluid basis, and obstacle position and dimension, were examined to identify optimal conditions for enhancing heat transfer quality. Principal findings indicated that increasing the Rayleigh number boosts buoyancy forces and alters vortex More > Graphic Abstract

Toni Dwi Novianto^1,2, Sri Rahayoe^1,*, Bakti Berlyanto Sedayu^2,*

Journal of Renewable Materials, Vol.13, No.2, pp. 385-400, 2025, DOI:10.32604/jrm.2024.057313 - 20 February 2025

Abstract This study aimed to develop and characterize biodegradable packaging film from blends of two natural polysaccharides, i.e., agar and glucomannan. The glucomannan used was derived from the specific tuber plant Amorphophallus oncophyllus (locally known as “porang”), which grows abundantly in Indonesian forests and remains underutilized. Various ratios of agar and porang-glucomannan (PG) proportions were formulated to produce a food packaging film, which was subsequently tested for its mechanical, physical, chemical, and thermal properties. The results showed that the inclusion of PG to the film formulations notably enhanced the stretchability of agar films, achieving maximum a… More > Graphic Abstract

Jinxing Li¹, Mustafa Zor², Xiaojian Zhou³, Guanben Du³, Denis Rodrigue⁴, Xiaodong (Alice) Wang^1,*

Journal of Renewable Materials, Vol.13, No.2, pp. 329-347, 2025, DOI:10.32604/jrm.2024.056795 - 20 February 2025

Abstract In this study, two series of foams based on tannic acid (TA), furfuryl alcohol (FA), soybean protein isolate (SPI), and casein (CA), namely TA–FA–SPI (TS series) and TA–FA–CA (TC series) were developed, and their properties were enhanced by adding poplar fibers (WF). From the samples produced, a complete set of characterization was performed including possible crosslinking reactions, morphology, mechanical properties, flame retardancy, thermal insulation and thermal stability. Fourier-transform infrared spectroscopy (FTIR) revealed possible covalent crosslinking among the components and hydrogen bonding between WF and the matrix. Viscosity results indicated that lower prepolymer viscosity led to… More >

Hong Lei¹, Xiaojian Zhou², Antonio Pizzi^3,*, Guanben Du^2,*, Xuedong Xi²

Journal of Renewable Materials, Vol.13, No.2, pp. 199-249, 2025, DOI:10.32604/jrm.2025.02024-0048 - 20 February 2025

Abstract This review is composed of three main parts each of which is written by well-known top specialists that have been, in a way or other, also the main participants of the majority of the developments reported. Thus, after a general part covering the grand lines and more in-depth views of more recent tannin, lignin, carbohydrate and soy bioadhesives, some mix of the other bio raw materials with soy protein and soy flour and some other differently sourced bioadhesives for wood, this review presents a more in-depth part on starch-based wood adhesives and a more in-depth… More > Graphic Abstract

Song Liu^1,2, Hongying Zhang^1,*, Xue Li¹, Xi Yang^1,3

CMC-Computers, Materials & Continua, Vol.82, No.2, pp. 1969-1984, 2025, DOI:10.32604/cmc.2024.057662 - 17 February 2025

Abstract Enhancing low-light images with color distortion and uneven multi-light source distribution presents challenges. Most advanced methods for low-light image enhancement are based on the Retinex model using deep learning. Retinexformer introduces channel self-attention mechanisms in the IG-MSA. However, it fails to effectively capture long-range spatial dependencies, leaving room for improvement. Based on the Retinexformer deep learning framework, we designed the Retinexformer+ network. The “+” signifies our advancements in extracting long-range spatial dependencies. We introduced multi-scale dilated convolutions in illumination estimation to expand the receptive field. These convolutions effectively capture the weakening semantic dependency between pixels… More >