相关参考文献
注意：仅列出部分参考文献，下载原文获取全部文献信息。
Article
Optics
Yanlong Cao et al.
Summary: Visual surface defect inspection is important for assessing product quality in various industrial applications, but it remains challenging to develop deep-learning-based approaches for industrial product defect inspection. To overcome this challenge, we propose a light-weight defect classification model based on pre-trained SqueezeNet architecture and introduce three effective techniques to achieve high accuracy with limited defective training samples. Experimental results validate the effectiveness of our method.
OPTICS AND LASERS IN ENGINEERING
(2023)
Review
Materials Science, Multidisciplinary
Xin Wen et al.
Summary: Steel surface defect recognition is important in industrial defect detection and has gained increasing attention. This paper discusses the key hardware and options for steel surface defect detection systems, and provides a literature review of algorithms for steel surface defect recognition, including traditional machine learning algorithms based on texture and shape features, as well as supervised, unsupervised, and weakly supervised deep learning algorithms. Common datasets and algorithm performance evaluation metrics in this field are also summarized. Lastly, the challenges and corresponding solutions for current steel surface defect recognition algorithms are discussed, along with the future work focus.
Article
Automation & Control Systems
Minghui Yang et al.
Summary: This study proposes an end-to-end memory-based segmentation network (MemSeg) for high-accuracy and real-time semi-supervised image surface defect detection in industrial scenarios. By introducing artificially simulated abnormal samples and memory samples, MemSeg achieves a good balance between accuracy and speed. It explicitly learns the differences between normal and simulated abnormal images, and uses a memory pool to store general patterns of normal samples for effective guesses about abnormal regions.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Materials Science, Textiles
Hongwei Zhang et al.
Summary: In this article, an Attention-based Feature Fusion Generative Adversarial Network framework is proposed for the unsupervised defect detection of yarn-dyed fabrics. The framework utilizes a modified Feature Pyramid Network to fuse multi-level information and an attention mechanism to enhance feature representation capabilities. Experimental results demonstrate the effectiveness of the proposed method compared to other methods on public datasets.
TEXTILE RESEARCH JOURNAL
(2023)
Article
Materials Science, Textiles
Hongwei Zhang et al.
Summary: This article introduces an Attention-Gate-based U-shaped Reconstruction Network (AGUR-Net) and a dual-threshold segmentation post-processing method for unsupervised defect detection in color-patterned fabrics. AGUR-Net consists of an encoder, an Atrous Spatial Pyramid Pooling module, and an attention gate weighted fusion residual decoder, which can better detect and locate defects in color-patterned fabrics.
TEXTILE RESEARCH JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Liyang Chen et al.
Summary: This paper introduces UTRAD, a U-Transformer based Anomaly Detection framework that achieves a more stable training process and precise anomaly detection and localization results by reconstructing more informative feature distribution instead of raw images. UTRAD also reduces computational cost and memory usage with a multi-scale pyramidal hierarchy and skip connections.
Article
Computer Science, Artificial Intelligence
Jie Yang et al.
Summary: In this paper, an unsupervised mechanism called deep feature correspondence (DFC) is proposed to detect and segment anomalies in images using prior knowledge from anomaly-free samples. The DFC utilizes an asymmetric dual network framework and incorporates self-feature enhancement and multi-context residual learning modules to improve detection performance. The approach achieves state-of-the-art results on the benchmark dataset and outperforms other methods in real industrial inspection scenarios.
PATTERN RECOGNITION
(2022)
Article
Construction & Building Technology
Yanyan Wang et al.
Summary: Current defect detection studies in the industrial fields primarily rely on supervised strategies, which require a large amount of annotated defective samples. However, it is difficult to meet such data requirements in actual industrial scenarios. To address this issue, this paper proposes a novel approach for defect detection using only non-defective samples, which can accurately detect defects with complex backgrounds and weak textures.
AUTOMATION IN CONSTRUCTION
(2022)
Article
Computer Science, Information Systems
Sungwook Lee et al.
Summary: Previous studies on anomaly localization lacked adaptation to the target dataset. This paper proposes a Coupled-hypersphere-based Feature Adaptation method that accurately discriminates between normal and abnormal features, outperforming previous methods in both quantitative and qualitative evaluations. The paper also highlights the issue of biased features in pre-trained CNNs.
Article
Computer Science, Artificial Intelligence
Vitjan Zavrtanik et al.
Summary: The article discusses the issue of visual anomaly detection and the shortcomings of the auto-encoder method in reconstructing anomalous regions. A new approach (RIAD) is proposed to address this problem by reconstructing images through inpainting, setting a new state-of-the-art in anomaly detection benchmarks.
PATTERN RECOGNITION
(2021)
Article
Engineering, Multidisciplinary
Yanyan Wang et al.
Summary: This article proposes a new method called RENet for accurate and robust pavement crack detection, utilizing a rectangular convolution pyramid and edge enhancement network. Experimental results demonstrate that this method surpasses other state-of-the-art algorithms in terms of robustness and universality.
Article
Optics
Jie Liu et al.
Summary: In this research, a semi-supervised anomaly detection method based on Dual Prototype Auto-Encoder is proposed, which achieves effective results by introducing dual prototype loss and reconstruction loss to encourage latent vectors to stay closer.
OPTICS AND LASERS IN ENGINEERING
(2021)
Article
Optics
Defu Zhang et al.
OPTICS AND LASERS IN ENGINEERING
(2020)
Article
Automation & Control Systems
Xian Tao et al.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2020)
Article
Engineering, Electrical & Electronic
Ling Xiao et al.
IEEE SENSORS JOURNAL
(2020)
Article
Computer Science, Hardware & Architecture
Ian Goodfellow et al.
COMMUNICATIONS OF THE ACM
(2020)
Article
Optics
Yu He et al.
OPTICS AND LASERS IN ENGINEERING
(2019)
Article
Computer Science, Artificial Intelligence
K. L. Mak et al.
IMAGE AND VISION COMPUTING
(2009)
