Jianfeng Liu¹, Yongjiao Yang¹, Kangyi Yang¹, Changhua Hu¹, Zijia Xu¹, Qingguo Shi², Yi Su^2,*

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.082804 - 15 June 2026

Abstract Power enterprise inspection and supervision require greater intelligence, efficiency, and standardization; however, existing approaches are limited by inefficient knowledge retrieval, inaccurate issue identification, and insufficient support for standardized reporting and rectification tracking. This study proposes a lightweight, domain-adaptive large language model (LLM) framework based on Low-Rank Adaptation (LoRA), integrating Retrieval-Augmented Generation (RAG) and structured prompt engineering to enable evidence-grounded inspection tasks. The framework achieves parameter-efficient adaptation through low-rank decomposition and constructs a domain-specific multimodal knowledge base, enhancing output traceability, consistency, and task generalization. A key contribution is the introduction of a Sensitive Information Control Gate, More >

Haotian Cao^1,2, Qingsheng Zhu^1,2,3,*

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.081695 - 15 June 2026

Abstract Point Cloud Registration (PCR) is a basic task in computer vision, mobile robotics, and autonomous driving. PCR primarily faces challenges, including insufficient registration performance in low-overlap scenarios and high computational resource consumption in large-scale point cloud scenarios. Most recent PCR methods are transformer-based. Methods like transformers have quadratic computational complexity $𝒪(n^{2}d)$, leading to rapid increases in computational cost with large-scale point cloud data. To address these problems, an iterative PCR method named Attention and Mamba Based Iterative Registration Network (AMBIR) is proposed, overcoming the shortcomings of the current PCR method on low-overlap and large-scale scenarios. Specifically, an… More >

Zhichao Pei, Ou Ye^*, Panyu Yang, Kaiwen He

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.081490 - 15 June 2026

Abstract To address the issue of insufficient transferability of existing adversarial example generation methods for vision-language pre-training (VLP) models, this paper proposes an adversarial example transfer method for VLP models based on negative sample feature perturbation. First, a novel cross-modal collaborative perturbation strategy is constructed. By introducing negative samples into the cross-modal perturbation mechanism, the strategy explores more perturbation directions, breaks the original modal alignment constraints and avoids the local focus of adversarial perturbations. Then, to reduce the computational cost, a dynamic threshold attack strategy is built to measure the modal similarity of the generated adversarial… More >

Junyi Wang¹, Jianghai Geng^1,*, Jiaqi Liu², Haibin Zhu³

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.080895 - 15 June 2026

Abstract In intelligent manufacturing and remanufacturing systems, the thermal safety of the power distribution infrastructure is crucial for ensuring production continuity, equipment reliability, and operational resilience. Traditional temperature monitoring methods often have problems such as high deployment costs, strong environmental sensitivity, or limited physical interpretability in distributed workshop environments. To address these limitations, this study proposes a physically information-driven intelligent thermal color-changing fault identification framework. Based on thermochromic experiments, irreversible color-changing coatings are selected, which are combined with a visual-based computing pipeline for autonomous overheating detection. The framework proposes a thermal fault temperature identification algorithm based… More >

Langyue Zhao^1,2, Yubin Yuan^3,*, Yiquan Wu^2,*

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.080232 - 15 June 2026

Abstract This paper presents a systematic survey of machine vision-based surface defect detection technologies, focusing on five core challenges in the field: interference from complex backgrounds, small object detection, class imbalance, dynamic scene modeling, and cross-scenario generalization. It reviews key technical approaches corresponding to these challenges over the past five years. Furthermore, a dataset characterization analysis framework is established around these challenges, summarizing and comparing the characteristics of over 40 publicly available datasets across more than ten scenarios, including PCB, photovoltaic, metal, and pavement surfaces. Quantitative selection metrics (such as the small target coefficient and texture More >

Ying Tang¹, Chuanyi Ma², Feng Guo^1,*, Wenhao Sun¹

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.079851 - 15 June 2026

Abstract Road Abandoned Objects (RAOs) pose significant threats to traffic safety, particularly due to their small size, irregular shapes, and unpredictable distribution in complex road environments. The primary objective of this study is to develop an accurate and real-time detection framework for RAOs while maintaining low computational cost for practical deployment. To achieve this, we propose RAO-YOLO, a lightweight vision-based detection framework built upon an enhanced YOLO architecture. Specifically, a Mixed Aggregation Network (MANet) is introduced to improve multi-scale feature representation, and a Lightweight Shared Detail-Enhanced Detection (LSDD) head is designed to enhance localization accuracy for More >

Ramakrishna Gandi^1,*, Geetha A.¹, Ramasubbareddy B.²

CMC-Computers, Materials & Continua, Vol.88, No.2, 2026, DOI:10.32604/cmc.2026.077569 - 15 June 2026

Abstract In the fields of affective computing, human-computer interaction, and psychological evaluation, the capacity to recognize emotions is crucial. Unimodal systems in the form of visual systems or of the physiological type are usually not designed to capture the complexity that exists in emotional states. The paper proposes NeuroVision: Multimodal Emotion Recognition System, combining facial video frames information and electroencephalogram (EEG) based information to enhance the accuracy and stability of the system. The system applies ResNet50 on the spatial information of facial expressions, Vision Transformer (ViT) on the temporal movements in the video, and an EEG-MLP… More >

Bo Yang^1,*, Chao Liu²

CMES-Computer Modeling in Engineering & Sciences, Vol.147, No.2, 2026, DOI:10.32604/cmes.2026.083726 - 27 May 2026

Abstract This article has no abstract. More >

Chuanjiang Chen¹, Junyong Zhou^1,*, Binbin Du¹, Miaosi Dong^2,*, Liwen Zhang¹, Bitang Zhu³

CMES-Computer Modeling in Engineering & Sciences, Vol.147, No.2, 2026, DOI:10.32604/cmes.2026.081546 - 27 May 2026

Abstract Computer vision has been widely adopted in intelligent construction monitoring; however, existing studies primarily focus on identifying individual construction elements or isolated activities, with limited capability for integrated monitoring of complete construction workflows. Such workflow-level automation is a prerequisite for intelligent construction and unmanned job sites. To address the challenge of reliable visual recognition in drill-and-blast tunnel environments characterized by uneven illumination, localized glare, and dust interference, this study proposes a methodological framework for construction workflow recognition at the tunnel face using computer vision and context reasoning. The framework consists of three components: (1) a… More >

Munthir Qasaimeh¹, Mostafa Ali¹, Qasem Abu Al-Haija^2,*

CMES-Computer Modeling in Engineering & Sciences, Vol.147, No.2, 2026, DOI:10.32604/cmes.2026.080864 - 27 May 2026

Abstract Vision Transformers (ViTs) have recently achieved high performance in retinal Optical Coherence Tomography (OCT) classification studies. However, ViT models continue to face significant challenges, including high computational cost, vulnerability to adversarial attacks, and pronounced sensitivity to preprocessing techniques. This study introduces GreenShield, a unified framework designed to produce an efficient and robust ViT model, referred to as GreenShield-ViT, which outperforms existing lightweight ViT variants in terms of adversarial robustness for retinal OCT classification. The framework integrates a gradient-based block-importance pruning strategy to compress the ViT/B-16 architecture, and adversarial training with proper ImageNet normalization and anti-saturation… More >