Min Yao^1,*, Guangjie Hu¹, Yaozu Zhang²

CMC-Computers, Materials & Continua, Vol.83, No.2, pp. 2751-2771, 2025, DOI:10.32604/cmc.2025.060860 - 16 April 2025

Abstract Semantic segmentation of remote sensing images is a critical research area in the field of remote sensing. Despite the success of Convolutional Neural Networks (CNNs), they often fail to capture inter-layer feature relationships and fully leverage contextual information, leading to the loss of important details. Additionally, due to significant intra-class variation and small inter-class differences in remote sensing images, CNNs may experience class confusion. To address these issues, we propose a novel Category-Guided Feature Collaborative Learning Network (CG-FCLNet), which enables fine-grained feature extraction and adaptive fusion. Specifically, we design a Feature Collaborative Learning Module (FCLM)… More >

Pratik Jadhav¹, Vuppala Adithya Sairam¹, Niranjan Bhojane¹, Abhyuday Singh¹, Shilpa Gite^1,2, Biswajeet Pradhan^3,*, Mrinal Bachute¹, Abdullah Alamri⁴

CMC-Computers, Materials & Continua, Vol.83, No.2, pp. 3493-3517, 2025, DOI:10.32604/cmc.2025.060764 - 16 April 2025

Abstract Electronic nose and thermal images are effective ways to diagnose the presence of gases in real-time real-time. Multimodal fusion of these modalities can result in the development of highly accurate diagnostic systems. The low-cost thermal imaging software produces low-resolution thermal images in grayscale format, hence necessitating methods for improving the resolution and colorizing the images. The objective of this paper is to develop and train a super-resolution generative adversarial network for improving the resolution of the thermal images, followed by a sparse autoencoder for colorization of thermal images and a multimodal convolutional neural network for… More >

Jianxin Feng^1,2,*, Xiaoyao Liu^1,2

CMC-Computers, Materials & Continua, Vol.83, No.2, pp. 2087-2107, 2025, DOI:10.32604/cmc.2025.060252 - 16 April 2025

Abstract As a form of discrete representation learning, Vector Quantized Variational Autoencoders (VQ-VAE) have increasingly been applied to generative and multimodal tasks due to their ease of embedding and representative capacity. However, existing VQ-VAEs often perform quantization in the spatial domain, ignoring global structural information and potentially suffering from codebook collapse and information coupling issues. This paper proposes a frequency quantized variational autoencoder (FQ-VAE) to address these issues. The proposed method transforms image features into linear combinations in the frequency domain using a 2D fast Fourier transform (2D-FFT) and performs adaptive quantization on these frequency components… More >

Chunlai Du^1,#, Xin Gu^1,#, Yanhui Guo^2,*, Siqi Guo³, Ziwei Pang³, Yi Du³, Guoqing Du^3,*

CMC-Computers, Materials & Continua, Vol.83, No.2, pp. 3435-3450, 2025, DOI:10.32604/cmc.2025.060212 - 16 April 2025

Abstract Coronary artery disease is a highly lethal cardiovascular condition, making early diagnosis crucial for patients. Echocardiograph is employed to identify coronary heart disease (CHD). However, due to issues such as fuzzy object boundaries, complex tissue structures, and motion artifacts in ultrasound images, it is challenging to detect CHD accurately. This paper proposes an improved Transformer model based on the Feedback Self-Attention Mechanism (FSAM) for classification of ultrasound images. The model enhances attention weights, making it easier to capture complex features. Experimental results show that the proposed method achieves high levels of accuracy, recall, precision, F1 More >

Xinyi Song¹, Haixiang Zhang^1,*, Shaohua Li²

CMC-Computers, Materials & Continua, Vol.83, No.2, pp. 2735-2750, 2025, DOI:10.32604/cmc.2025.059006 - 16 April 2025

Abstract In recent years, Transformer has achieved remarkable results in the field of computer vision, with its built-in attention layers effectively modeling global dependencies in images by transforming image features into token forms. However, Transformers often face high computational costs when processing large-scale image data, which limits their feasibility in real-time applications. To address this issue, we propose Token Masked Pose Transformers (TMPose), constructing an efficient Transformer network for pose estimation. This network applies semantic-level masking to tokens and employs three different masking strategies to optimize model performance, aiming to reduce computational complexity. Experimental results show More >

Abdel Aziz Osseni¹, Hubert Olivier Dossou-Yovo^2,*, Apollon D.M.T. Hegbe³, Muhammad Nauman Khan⁴, Brice Sinsin²

Revue Internationale de Géomatique, Vol.34, pp. 209-234, 2025, DOI:10.32604/rig.2025.061448 - 14 April 2025

Abstract Today, environmental studies based on satellite imagery are known as making valuable contributions to the dynamics and spatial prediction of sensitive or complex ecosystems such as wide protected areas and represent sustainable decision tools. The Pendjari and W Transboundary Reserves which constitute biodiversity reservoirs, habitats for wildlife conservation lack substantial investigations on the vegetation dynamics. Despite the protection measures they benefit from, these reserves remain dependent on climatic hazards that can influence their stability. The present study is innovative since it applied remote sensing techniques combined with climate records from the last thirty years to… More >

Mohamed Medani¹, Yahia Said², Nashwan Adnan Othman^3,4, Farrukh Yuldashev⁵, Mohamed Kchaou⁶, Faisal Khaled Aldawood⁶, Bacha Rehman^7,*

CMES-Computer Modeling in Engineering & Sciences, Vol.143, No.1, pp. 1049-1070, 2025, DOI:10.32604/cmes.2025.063405 - 11 April 2025

Abstract Satellite images are widely used for remote sensing and defence applications, however, they are subject to a variety of threats. To ensure the security and privacy of these images, they must be watermarked and encrypted before communication. Therefore, this paper proposes a novel watermarked satellite image encryption scheme based on chaos, Deoxyribonucleic Acid (DNA) sequence, and hash algorithm. The watermark image, DNA sequence, and plaintext image are passed through the Secure Hash Algorithm (SHA-512) to compute the initial condition (keys) for the Tangent-Delay Ellipse Reflecting Cavity Map (TD-ERCS), Henon, and Duffing chaotic maps, respectively. Through More >

Jahanzaib Latif ¹, Ahsan Wajahat¹, Alishba Tahir², Anas Bilal^3,*, Mohammed Zakariah⁴, Abeer Alnuaim⁴

CMES-Computer Modeling in Engineering & Sciences, Vol.143, No.1, pp. 539-567, 2025, DOI:10.32604/cmes.2025.062301 - 11 April 2025

Abstract Glaucoma, a leading cause of blindness, demands early detection for effective management. While AI-based diagnostic systems are gaining traction, their performance is often limited by challenges such as varying image backgrounds, pixel intensity inconsistencies, and object size variations. To address these limitations, we introduce an innovative, nature-inspired machine learning framework combining feature excitation-based dense segmentation networks (FEDS-Net) and an enhanced gray wolf optimization-supported support vector machine (IGWO-SVM). This dual-stage approach begins with FEDS-Net, which utilizes a fuzzy integral (FI) technique to accurately segment the optic cup (OC) and optic disk (OD) from retinal images, even More >

Mohammad Barr^*

CMES-Computer Modeling in Engineering & Sciences, Vol.143, No.1, pp. 593-616, 2025, DOI:10.32604/cmes.2025.062264 - 11 April 2025

Abstract Recent years have seen a surge in interest in object detection on remote sensing images for applications such as surveillance and management. However, challenges like small object detection, scale variation, and the presence of closely packed objects in these images hinder accurate detection. Additionally, the motion blur effect further complicates the identification of such objects. To address these issues, we propose enhanced YOLOv9 with a transformer head (YOLOv9-TH). The model introduces an additional prediction head for detecting objects of varying sizes and swaps the original prediction heads for transformer heads to leverage self-attention mechanisms. We… More >

Abdullah A. Asiri¹, Toufique A. Soomro^2,3,*, Ahmed Ali⁴, Faisal Bin Ubaid⁵, Muhammad Irfan^6,*, Khlood M. Mehdar⁷, Magbool Alelyani⁸, Mohammed S. Alshuhri⁹, Ahmad Joman Alghamdi¹⁰, Sultan Alamri¹⁰

CMES-Computer Modeling in Engineering & Sciences, Vol.143, No.1, pp. 255-287, 2025, DOI:10.32604/cmes.2025.061683 - 11 April 2025

Abstract Global mortality rates are greatly impacted by malignancies of the brain and nervous system. Although, Magnetic Resonance Imaging (MRI) plays a pivotal role in detecting brain tumors; however, manual assessment is time-consuming and susceptible to human error. To address this, we introduce ICA2-SVM, an advanced computational framework integrating Independent Component Analysis Architecture-2 (ICA2) and Support Vector Machine (SVM) for automated tumor segmentation and classification. ICA2 is utilized for image preprocessing and optimization, enhancing MRI consistency and contrast. The Fast-Marching Method (FMM) is employed to delineate tumor regions, followed by SVM for precise classification. Validation on More >