Related references
Note: Only part of the references are listed.
Article
Computer Science, Artificial Intelligence
Kai Han et al.
Summary: Transformer, a deep neural network with a self-attention mechanism, has been initially used in natural language processing and is now gaining attention in computer vision tasks. Transformer-based models perform as well as or even better than convolutional and recurrent neural networks in various visual benchmarks. This paper reviews vision transformer models, categorizes them based on different tasks, and analyzes their advantages and disadvantages. The discussed categories include backbone network, high/mid-level vision, low-level vision, and video processing. Efficient methods for applying transformer in real device-based applications are also explored. The challenges and further research directions for vision transformers are discussed as well.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Construction & Building Technology
Erhu Zhang et al.
Summary: Cracks pose a serious threat to the safety of hydraulic dams, but timely detection remains a challenge. To address this, we propose a pixel-level crack detection network, UTCD-Net, which combines the transformer and CNN models. Our method captures both local and global crack features, enabling the detection of thin and long cracks.
AUTOMATION IN CONSTRUCTION
(2023)
Article
Construction & Building Technology
Qiwen Qiu et al.
Summary: This paper proposes the integration of You Only Look Once (YOLO) into an unmanned aerial vehicle (UAV) for real-time crack detection in tiled sidewalks. Different network architectures of YOLOv2-tiny, Darknet19-based YOLOv2, ResNet50-based YOLOv2, YOLOv3, and YOLOv4-tiny are compared to improve accuracy and speed of detection. The results show that ResNet50-based YOLOv2 and YOLOv4-tiny offer excellent accuracy and speed, and remarkable ability in detecting small cracks. They also demonstrate good adaptability to environmental conditions such as shadows, rain, and motion-induced blurriness. The evaluation suggests the appropriate altitude and scanning area for the YOLO-UAV-based platform to achieve remote, reliable, and rapid crack detection.
AUTOMATION IN CONSTRUCTION
(2023)
Article
Computer Science, Interdisciplinary Applications
Zhong Zhou et al.
Summary: In this paper, a hybrid semantic segmentation algorithm named SCDeepLab is proposed, which addresses the limitations of CNN-based algorithms in capturing global features and the weaknesses of Transformer-based networks in losing local features. SCDeepLab combines the advantages of Swin Transformer and CNN to effectively extract both global semantic and local detailed information, resulting in higher segmentation accuracy compared to previous methods.
COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING
(2023)
Article
Computer Science, Interdisciplinary Applications
Pengcheng Jiao et al.
Summary: This study develops a real-time marine and offshore structural health monitoring system based on controllable underwater robots. It includes three modules: underwater monitoring robots, vision-based image processing and analyzing, and time-dependent damage assessing and early warning. The system provides design guidance for next-generation multifunctional underwater devices.
COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING
(2023)
Article
Engineering, Multidisciplinary
Jingtao Zhong et al.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Feiniu Yuan et al.
Summary: This paper proposes a CNN and Transformer Complementary Network (CTC-Net) for medical image segmentation. It designs two encoders to produce complementary features in Transformer and CNN domains, and uses cross-domain fusion, feature correlation, and dual attention to enhance the representation ability of features. It also incorporates skip connections to extract spatial details, contextual semantics, and long-range information.
PATTERN RECOGNITION
(2023)
Article
Construction & Building Technology
Chao Xiang et al.
Summary: This study proposes a dual-encoder network called DTrC-Net, which combines transformers and convolutional neural networks, to address the challenges in crack segmentation caused by complex scenes. The DTrC-Net captures both local features and global contextual information of crack images and enhances feature fusion between adjacent and codec layers. Experimental results show that DTrC-Net achieves better predictions than other segmentation networks, with high accuracy (75.60%), recall (78.86%), F1-score (76.44%), and intersection over union (64.30%) on the Crack3238 dataset. It also achieves a fast processing speed of 78 frames per second.
AUTOMATION IN CONSTRUCTION
(2023)
Article
Engineering, Biomedical
Shengyong Diao et al.
Summary: Age-related macular degeneration (AMD) is a common cause of visual impairment in the elderly, characterized by drusen and choroidal neovascularization (CNV). This paper proposes a deep learning framework that utilizes dual guidance for image classification and segmentation tasks in AMD diagnosis. The framework achieves higher accuracy in classification and segmentation compared to other networks tested on public datasets. The results also demonstrate the generalizability of the proposed model for detecting macular edema and segmentation of retinal fluids.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Computer Science, Artificial Intelligence
Huisi Wu et al.
Summary: The study introduces a novel skin lesion segmentation method named FAT-Net, which integrates transformer branch and feature adaptation module to capture long-range dependencies and enhance feature fusion. Experimental results demonstrate the superior accuracy and inference speed of FAT-Net on four public datasets compared to state-of-the-art methods.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Engineering, Electrical & Electronic
Jie Pang et al.
Summary: In this paper, a novel Deep Crack Segmentation Network (DcsNet) is proposed, which achieves a balance of speed and accuracy through two feature extraction branches. Extensive experiments show that the proposed network achieves a good trade-off between accuracy and inference speed, outperforming state-of-the-art methods.
SIGNAL IMAGE AND VIDEO PROCESSING
(2022)
Review
Construction & Building Technology
Zhiguo He et al.
Summary: Integrated structural health monitoring ensures the functionality and operation of bridges through mechanism analysis, monitoring technology, and data analytics. This review discusses the current process and future trends of bridge monitoring, focusing on cutting-edge SHM technologies, data transmission and analytics methods, and prediction and early-warning models.
AUTOMATION IN CONSTRUCTION
(2022)
Article
Construction & Building Technology
Pai Pan et al.
Summary: This study focuses on crack detection on the outer surface of a nuclear containment in order to ensure nuclear power plant safety. A semantic segmentation model based on multi-feature fusion and focal loss is proposed to improve the crack segmentation performance. Comparative experiments and generalization experiments prove that the proposed method performs better than other commonly used methods.
CONSTRUCTION AND BUILDING MATERIALS
(2022)
Article
Engineering, Electrical & Electronic
Zhong Qu et al.
Summary: This paper proposes a method called CrackT-net, which uses convolutional neural networks (CNN) and Transformer to solve the problem of automatic crack segmentation. In the network design, a new backbone network RF UNet++ is used to enhance feature representation capabilities, and Transformer is used to capture more long dependencies and global context information. The effectiveness of the proposed method is demonstrated through evaluation on public datasets.
JOURNAL OF ELECTRONIC IMAGING
(2022)
Article
Construction & Building Technology
Wenjun Wang et al.
Summary: In this study, a novel SegCrack model for pixel-level crack segmentation using deep learning methods is proposed. The model utilizes a hierarchically structured Transformer encoder to output multiscale features and incorporates a top-down pathway and lateral connections for progressive feature upsampling and fusion. An online hard example mining strategy is also adopted to improve model performance. Experimental results demonstrate SegCrack achieves high precision, recall, F1 score, and mean intersection over union on the test set.
AUTOMATION IN CONSTRUCTION
(2022)
Article
Computer Science, Artificial Intelligence
Jianming Zhang et al.
Summary: The study introduces a novel traffic sign detection network named ReYOLO, which efficiently detects small and ambiguous traffic signs in the wild by learning rich contextual information and sensing scale variations. By using structural reparameterization methods and a novel weighting mechanism, the model is able to learn more effective features and narrow the semantic gap between multiple scales.
JOURNAL OF AMBIENT INTELLIGENCE AND SMART ENVIRONMENTS
(2022)
Article
Automation & Control Systems
Jian Zhang et al.
Summary: This research proposes an automatic detection and segmentation method for bridge surface cracks based on computer vision deep learning models, which is able to effectively identify and segment bridge cracks. Experimental results demonstrate that our method outperforms other baseline methods, with smaller model size and higher frame per second (FPS) performance.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Article
Engineering, Multidisciplinary
Yidan Yan et al.
Summary: This paper proposes a novel segmentation network for crack image detection under low-light conditions. A cycle generative adversarial network is used to translate low-light images to the bright domain, and an encoder-decoder segmentation network is employed for final crack detection. Experimental results show that the proposed network performs superiorly in both low-light and well-light conditions, indicating its potential for inspection tasks in poor lighting environments.
Article
Chemistry, Multidisciplinary
Meng Wu et al.
Summary: This paper introduces a new U-shaped convolutional neural network called TMCrack-Net for crack information detection in mural conservation. The authors propose a new network structure and incorporate feature pyramids and Transformer to optimize feature extraction and fusion, addressing the issues of current mainstream networks in crack detection. Experimental results demonstrate the superior performance of this method in crack detection.
APPLIED SCIENCES-BASEL
(2022)
Article
Construction & Building Technology
L. Minh Dang et al.
Summary: This research focuses on implementing computer vision techniques and deep learning to automate crack segmentation and real-life crack length measurement of masonry walls. The experimental results demonstrate that deep learning-based crack segmentation outperforms previous approaches and can provide accurate measurements.
CONSTRUCTION AND BUILDING MATERIALS
(2022)
Article
Computer Science, Artificial Intelligence
Jingdong Wang et al.
Summary: The High-Resolution Network (HRNet) maintains high-resolution representations and exchanges information across resolutions, resulting in superior performance in various applications such as human pose estimation, semantic segmentation, and object detection.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Article
Construction & Building Technology
Wenjun Wang et al.
Summary: This paper proposes a semi-supervised semantic segmentation network for crack detection, which reduces the requirement of annotated data and improves the model accuracy through the collaboration of student and teacher models. It can reduce the annotation workload while maintaining high accuracy.
AUTOMATION IN CONSTRUCTION
(2021)
Article
Automation & Control Systems
Wenbo Jiang et al.
Summary: The study proposed HDCB-Net for pixel-level detection of blurred cracks, achieving efficient fast crack detection through a two-stage strategy with a processing time of only 0.64 seconds per image. Adding to that, a public dataset comprising 150,632 high-resolution images was established for crack detection research purposes.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Construction & Building Technology
Ankang Ji et al.
AUTOMATION IN CONSTRUCTION
(2020)
Article
Automation & Control Systems
Wooram Choi et al.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2020)
Article
Computer Science, Information Systems
Jianming Zhang et al.
Article
Computer Science, Artificial Intelligence
Qin Zou et al.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2019)
Article
Computer Science, Artificial Intelligence
Yahui Liu et al.
Article
Computer Science, Artificial Intelligence
Yun Liu et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2019)
Article
Chemistry, Multidisciplinary
Jianming Zhang et al.
APPLIED SCIENCES-BASEL
(2019)
Article
Computer Science, Artificial Intelligence
Vijay Badrinarayanan et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2017)
Article
Engineering, Civil
Yong Shi et al.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2016)
Article
Optics
Ahmed Mahgoub Ahmed Talab et al.
Article
Computer Science, Interdisciplinary Applications
L Abdel-Qader et al.
JOURNAL OF COMPUTING IN CIVIL ENGINEERING
(2003)
