☆ 4.7 Article

Efficient semi-supervised surface crack segmentation with small datasets based on consistency regularisation and pseudo-labelling

AUTOMATION IN CONSTRUCTION (2024)

Related references

Note: Only part of the references are listed.

Article Radiology, Nuclear Medicine & Medical Imaging

Evaluating semi-supervision methods for medical image segmentation: applications in cardiac magnetic resonance imaging

Sarah M. Hooper et al.

Summary: This study proposes a semi-supervised approach to train segmentation networks with limited labeled data, and evaluates the performance of four cardiac MR segmentation networks using heterogeneous datasets and clinical metrics. The results show strong agreement and generalization performance, and demonstrate that the model trained with only 100 labeled image slices can achieve comparable results to a model trained with over 16,000 labeled image slices.

JOURNAL OF MEDICAL IMAGING (2023)

Add to Collection

Article Engineering, Multidisciplinary

Semi-supervised learning framework for crack segmentation based on contrastive learning and cross pseudo supervision

Chao Xiang et al.

Summary: In this study, a novel semi-supervised deep learning framework, called SemiCrack, is proposed for crack segmentation. The framework combines contrastive learning and cross pseudo supervision (CPS) and utilizes pixel contrastive loss to improve segmentation accuracy. Experimental results demonstrate that SemiCrack outperforms many existing fully-supervised and semi-supervised segmentation algorithms on various publicly available datasets. The segmentation accuracy of TC-Net is higher than that of most fully-supervised segmentation networks, with an improvement of about 2% in Intersection of Union (IoU). Besides, SemiCrack achieves comparable accuracy to other fully-supervised algorithms with only 20% labeled data.

MEASUREMENT (2023)

Add to Collection

Review Construction & Building Technology

Structural crack detection using deep convolutional neural networks

Raza Ali et al.

Summary: This article reviews the application of Convolutional Neural Networks (CNN) in civil structure crack detection, emphasizing the importance of CNN in image classification and segmentation, as well as analyzing recent developments. In addition to discussing machine learning methods, it also introduces the limitations of manual processing and image processing techniques in crack detection.

AUTOMATION IN CONSTRUCTION (2022)

Add to Collection

Review Construction & Building Technology

A critical review and comparative study on image segmentation-based techniques for pavement crack detection

Narges Kheradmandi et al.

Summary: The article emphasizes the significance of early detection of pavement decay in road maintenance, introduces the application of image processing in this field, and highlights the challenges in pavement image segmentation. The literature review indicates the need for further improvement on existing research results, providing a direction and guidance for future studies.

CONSTRUCTION AND BUILDING MATERIALS (2022)

Add to Collection

Article Engineering, Multidisciplinary

Efficient attention-based deep encoder and decoder for automatic crack segmentation

Dong H. Kang et al.

Summary: In this paper, a novel semantic transformer representation network (STRNet) is developed for crack segmentation with fast processing speed and high performance. The network is trained and tested in complex scenes, achieving high precision, recall, F1 score, and mIoU. Comparing with other advanced networks, STRNet shows the best performance in evaluation metrics with the fastest processing speed.

STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL (2022)

Add to Collection

Article Green & Sustainable Science & Technology

Use of Parallel ResNet for High-Performance Pavement Crack Detection and Measurement

Zhun Fan et al.

Summary: In this study, the use of Convolutional Neural Networks (CNNs) for crack identification in road pavement health and safety assessment is explored. A novel deep residual convolutional neural network called Parallel ResNet is proposed to overcome the limitations of existing methods. Experimental results demonstrate that Parallel ResNet achieves high accuracy in crack detection and measurement, making it a reliable method for pavement crack image analysis.

SUSTAINABILITY (2022)

Add to Collection

Article Construction & Building Technology

Automated concrete crack evaluation using stereo vision with two different focal lengths

Hyunjun Kim et al.

Summary: This study introduces an advanced stereo vision framework using wide-angle and telephoto lenses for crack quantification and 3D reconstruction of concrete structures. A robust depth estimation strategy is also proposed and the performance was field validated.

AUTOMATION IN CONSTRUCTION (2022)

Add to Collection

Article Chemistry, Analytical

Automatic Detection of Cracks on Concrete Surfaces in the Presence of Shadows

Paulius Palevicius et al.

Summary: Deep learning-based methods, particularly convolutional neural networks, are used for automatic crack identification on concrete surfaces. However, these methods often overlook the complexity of the image collection process. This paper explores the challenges of using deep learning-based methods for concrete crack detection in the presence of shadows and proposes novel shadow augmentation techniques to improve accuracy.

SENSORS (2022)

Add to Collection

Article Construction & Building Technology

The development of a digitally enhanced visual inspection framework for masonry bridges in the UK

Saeed Talebi et al.

Summary: This research aims to develop a framework for digital visual inspections of masonry railway bridges using multiple NDT technologies to overcome the limitations of conventional inspection regimes. A digitally enhanced visual inspection framework is developed by using complementary optical methods, which require fewer subjective interpretations and can be conducted remotely, making it safer and more efficient compared to traditional methods.

CONSTRUCTION INNOVATION-ENGLAND (2022)

Add to Collection

Article Construction & Building Technology

Vision transformer-based autonomous crack detection on asphalt and concrete surfaces

Elyas Asadi Shamsabadi et al.

Summary: This research proposes a ViT-based framework for crack detection on asphalt and concrete surfaces, achieving enhanced real-world crack segmentation performance through transfer learning and IoU loss function. Compared to CNN-based models, TransUNet with a CNN-ViT backbone shows better average IoU on small and multiscale crack semantics and ViT helps the encoder-decoder network exhibit robust performance against various noisy signals.

AUTOMATION IN CONSTRUCTION (2022)

Add to Collection

Article Construction & Building Technology

TOPO-Loss for continuity-preserving crack detection using deep learning

B. G. Pantoja-Rosero et al.

Summary: We propose a method for segmenting cracks in images of masonry buildings damaged by earthquakes. The method combines multiple loss functions to improve the continuity of cracks and accommodate imprecise annotations. Experimental results show that combining these loss functions significantly improves the topology of crack predictions.

CONSTRUCTION AND BUILDING MATERIALS (2022)

Add to Collection

Article Engineering, Multidisciplinary

Robust crack detection in masonry structures with Transformers

Elyas Asadi Shamsabadi et al.

Summary: This paper explores the use of Transformer models for crack detection in masonry structures, demonstrating higher accuracy and better generalisation to noise compared to convolutional neural networks. TransUNet-hybrid achieves the best generalisation and robustness against noise in the experiments.

MEASUREMENT (2022)

Add to Collection

Article Engineering, Multidisciplinary

Experimental investigation on the use of multiple very low-cost inertial-based devices for comfort assessment and rail track monitoring

Rafael Henrique de Oliveira et al.

Summary: In this paper, the authors present the first experimental results on the use of multiple very low-cost sensors aboard trains for vibration monitoring. They propose a collective approach to provide more accurate and robust results. The potential of the collective approach is proven through noise reduction and discrepant sensor identification.

MEASUREMENT (2022)

Add to Collection

Article Construction & Building Technology

Attention-based generative adversarial network with internal damage segmentation using thermography

Rahmat Ali et al.

Summary: This paper introduces a real-time, high-performance deep-learning network for pixel-level segmentation of internal damages in concrete members using active thermography. By developing an attention-based generative adversarial network (AGAN) to train the proposed internal damage segmentation network (IDSNet), it achieved excellent performance on the test set.

AUTOMATION IN CONSTRUCTION (2022)

Add to Collection

Article Automation & Control Systems

Crack Segmentation Network using Additive Attention Gate-CSN-II

Raza Ali et al.

Summary: This study proposes a crack segmentation network architecture called CSN-II, which utilizes an additive attention gate to focus on significant crack regions in the image. Experimental results show that CSN-II achieves high accuracy on two crack image datasets using a local balanced cross-entropy loss function.

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE (2022)

Add to Collection

Article Engineering, Civil

Automatic detection of defects in concrete structures based on deep learning

Wenjun Wang et al.

Summary: The automated detection of concrete surface defects is a deep learning-based method that uses a one-stage object detection network, combining the backbone network EfficientNetB0 and feature pyramid network. Training and testing on the concrete surface defects dataset showed that the method achieved high detection accuracy.

STRUCTURES (2022)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Automatic railroad track components inspection using real-time instance segmentation

Feng Guo et al.

Summary: Regular track inspections in the United States heavily rely on manual inspections, which are low-efficiency and not as accurate as desired, especially for detecting missing and broken track components. To address this issue, a real-time pixel-level rail components detection framework has been proposed in this study, which achieved higher accuracy and real-time processing speed compared to existing models.

COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING (2021)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Semi-supervised learning based on convolutional neural network and uncertainty filter for facade defects classification

Jingjing Guo et al.

Summary: The study introduces a semi-supervised learning algorithm that trains an accurate classifier using only a small amount of labeled data, and improves classification accuracy by selecting reliable unlabeled data with an uncertainty filter. Validation experiments show that the proposed method significantly enhances model accuracy compared to traditional supervised learning algorithms, while also reducing the time and cost for preparing labeled data.

COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING (2021)

Add to Collection

Article Engineering, Multidisciplinary

A research on an improved Unet-based concrete crack detection algorithm

Lingxin Zhang et al.

Summary: The paper proposed the CrackUnet model based on deep learning, which uses resized, labeled, and augmented crack images to create a dataset, and adopts a new loss function called generalized Dice loss to improve crack detection accuracy. The study investigates the impact of dataset size and model depth on training time, detection accuracy, and speed, showing that the CrackUnet model outperforms other methods with strong robustness and generalization.

STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

A semi-supervised learning detection method for vision-based monitoring of construction sites by integrating teacher-student networks and data augmentation

Bo Xiao et al.

Summary: This research proposes a semi-supervised learning detection method for construction site monitoring using teacher-student networks and data augmentation, achieving high detection performance with a limited number of labeled data. Experimental results showed a mean average precision of 91% on the validation set while only requiring 30% of the training set.

ADVANCED ENGINEERING INFORMATICS (2021)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Automated Damage Localization and Quantification in Concrete Bridges Using Point Cloud-Based Surface-Fitting Strategy

Hyunjun Kim et al.

Summary: Digital image processing is used for automated damage evaluation, but quantifying surface damage in 3D space is difficult using only 2D images, and processing close-up images can be time-consuming. This study proposes a framework for automated damage evaluation using 3D laser scanning and employs a tailored fitting strategy to accurately identify surface damage.

JOURNAL OF COMPUTING IN CIVIL ENGINEERING (2021)

Add to Collection

Article Construction & Building Technology

Automatic crack classification and segmentation on masonry surfaces using convolutional neural networks and transfer learning

Dimitris Dais et al.

Summary: This study explores deep learning techniques for crack detection on masonry walls, achieving an accuracy of 95.3%. Leveraging transfer learning, crack segmentation on masonry surfaces at the pixel level was successfully implemented. This represents the first application of deep learning for pixel-level crack segmentation on masonry surfaces.

AUTOMATION IN CONSTRUCTION (2021)

Add to Collection

Article Computer Science, Information Systems

Deep Learning for Hemorrhagic Lesion Detection and Segmentation on Brain CT Images

Lu Li et al.

Summary: This paper proposes a U-net based deep learning framework for automatic detection and segmentation of hemorrhage strokes in CT brain images. By comparing different Deep Learning topologies, adopting adversarial training, and training and evaluating the model on two different datasets, the effectiveness, robustness, and advantages of the proposed deep learning model in hemorrhage lesion diagnosis have been demonstrated, making it possible to be a clinical decision support tool in stroke diagnosis.

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS (2021)

Add to Collection

Article Construction & Building Technology

Automatic segmentation of TBM muck images via a deep-learning approach to estimate the size and shape of rock chips

Xiaoxiong Zhou et al.

Summary: Real-time muck analysis is crucial for assisting tunnel boring machines in intelligent tunneling. A deep learning-based approach using a dual UNet with multi-scale inputs and side-output, along with a post-processing algorithm, was proposed to automatically segment muck images and estimate rock chip size and shape. Experimental results showed that this approach outperformed U-Net and conventional methods, nearly meeting the requirement for real-time TBM muck image analysis.

AUTOMATION IN CONSTRUCTION (2021)

Add to Collection

Article Construction & Building Technology

Semi-supervised learning with GAN for automatic defect detection from images

Gaowei Zhang et al.

Summary: This research introduces a semisupervised generative adversarial network (SSGAN) with two sub-networks for automatic defect detection, utilizing attention mechanism and dual loss functions to enhance segmentation performance and reduce data labeling efforts.

AUTOMATION IN CONSTRUCTION (2021)

Add to Collection

Article Construction & Building Technology

Semi-supervised semantic segmentation network for surface crack detection

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)

Add to Collection

Review Construction & Building Technology

Image-Based Crack Detection Methods: A Review

Hafiz Suliman Munawar et al.

Summary: Automated defect detection in infrastructure, utilizing image processing and machine learning techniques, is crucial for ensuring effectiveness and reliability. This paper reviews and compares image-based crack detection techniques, highlighting the most promising automated approaches for crack detection. A total of 30 research articles published in top tier journals and conferences in the past decade are analyzed comprehensively to provide insights into the advancements in this field.

INFRASTRUCTURES (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

A survey on semi-supervised learning

Jesper E. Van Engelen et al.

MACHINE LEARNING (2020)

Add to Collection

Article Construction & Building Technology

Image-based concrete crack detection in tunnels using deep fully convolutional networks

Yupeng Ren et al.

CONSTRUCTION AND BUILDING MATERIALS (2020)

Add to Collection

Article Engineering, Multidisciplinary

Automatic crack recognition for concrete bridges using a fully convolutional neural network and naive Bayes data fusion based on a visual detection system

Gang Li et al.

MEASUREMENT SCIENCE AND TECHNOLOGY (2020)

Add to Collection

Article Computer Science, Artificial Intelligence

Deep co-training for semi-supervised image segmentation

Jizong Peng et al.

PATTERN RECOGNITION (2020)

Add to Collection

Article Chemistry, Analytical

Automatic Tunnel Crack Detection Based on U-Net and a Convolutional Neural Network with Alternately Updated Clique

Gang Li et al.

SENSORS (2020)

Add to Collection

Article Engineering, Multidisciplinary

Multi-level feature fusion in densely connected deep-learning architecture and depth-first search for crack segmentation on images collected with smartphones

Qipei Mei et al.

STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL (2020)

Add to Collection

Article Engineering, Civil

Feature Pyramid and Hierarchical Boosting Network for Pavement Crack Detection

Fan Yang et al.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2020)

Add to Collection

Article Materials Science, Multidisciplinary

Ensemble of Deep Convolutional Neural Networks for Automatic Pavement Crack Detection and Measurement

Zhun Fan et al.

COATINGS (2020)

Add to Collection

Article Chemistry, Analytical

Automatic Pixel-Level Crack Detection on Dam Surface Using Deep Convolutional Network

Chuncheng Feng et al.

SENSORS (2020)

Add to Collection

Article Chemistry, Analytical

Automated Transverse Crack Mapping System with Optical Sensors and Big Data Analytics

Kwanghee Won et al.

SENSORS (2020)

Add to Collection

Article Chemistry, Analytical

Data-Driven Structural Health Monitoring and Damage Detection through Deep Learning: State-of-the-Art Review

Mohsen Azimi et al.

SENSORS (2020)

Add to Collection

Article Construction & Building Technology

Deep machine learning approach to develop a new asphalt pavement condition index

Hamed Majidifard et al.

CONSTRUCTION AND BUILDING MATERIALS (2020)

Add to Collection

Article Construction & Building Technology

Deep convolution neural network-based transfer learning method for civil infrastructure crack detection

Qiaoning Yang et al.

AUTOMATION IN CONSTRUCTION (2020)

Add to Collection

Article Construction & Building Technology

An integrated approach to automatic pixel-level crack detection and quantification of asphalt pavement

Ankang Ji et al.

AUTOMATION IN CONSTRUCTION (2020)

Add to Collection

Review Computer Science, Interdisciplinary Applications

Machine Learning for Crack Detection: Review and Model Performance Comparison

Yung-An Hsieh et al.

JOURNAL OF COMPUTING IN CIVIL ENGINEERING (2020)

Add to Collection

Article Construction & Building Technology

A spatial-channel hierarchical deep learning network for pixel-level automated crack detection

Yue Pan et al.

AUTOMATION IN CONSTRUCTION (2020)

Add to Collection

Article Computer Science, Information Systems

Multiscale and Adversarial Learning-Based Semi-Supervised Semantic Segmentation Approach for Crack Detection in Concrete Structures

Seungbo Shim et al.

IEEE ACCESS (2020)

Add to Collection

Article Computer Science, Information Systems

Semi-Supervised Semantic Segmentation Using Adversarial Learning for Pavement Crack Detection

Gang Li et al.

IEEE ACCESS (2020)

Add to Collection

Article Construction & Building Technology

Autonomous concrete crack detection using deep fully convolutional neural network

Cao Vu Dung et al.

AUTOMATION IN CONSTRUCTION (2019)

Add to Collection

Article Construction & Building Technology

Computer vision-based concrete crack detection using U-net fully convolutional networks

Zhenqing Liu et al.

AUTOMATION IN CONSTRUCTION (2019)

Add to Collection

Article Chemistry, Analytical

Use of Unmanned Aerial Vehicle Imagery and Deep Learning UNet to Extract Rice Lodging

Xin Zhao et al.

SENSORS (2019)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Robust Pixel-Level Crack Detection Using Deep Fully Convolutional Neural Networks

Mohamad Alipour et al.

JOURNAL OF COMPUTING IN CIVIL ENGINEERING (2019)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Automatic Pixel-Level Crack Detection and Measurement Using Fully Convolutional Network

Xincong Yang et al.

COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING (2018)

Add to Collection

Article Computer Science, Interdisciplinary Applications

Deep Learning-Based Crack Damage Detection Using Convolutional Neural Networks

Young-Jin Cha et al.

COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING (2017)

Add to Collection

Article Engineering, Civil

Automatic Road Crack Detection Using Random Structured Forests

Yong Shi et al.

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS (2016)

Add to Collection

Review Computer Science, Hardware & Architecture

A Few Useful Things to Know About Machine Learning

Pedro Domingos

COMMUNICATIONS OF THE ACM (2012)

Add to Collection

Article Computer Science, Artificial Intelligence

Crack Tree: Automatic crack detection from pavement images

Qin Zou et al.

PATTERN RECOGNITION LETTERS (2012)

Add to Collection

© Peeref 2019-2024. All rights reserved.