-
1دورية أكاديمية
المصدر: Journal of Big Data, Vol 10, Iss 1, Pp 1-24 (2023)
مصطلحات موضوعية: Metal surface defects, Lightweight convolutional encoder, Semi-supervised learning, Self-supervised learning, Computer engineering. Computer hardware, TK7885-7895, Information technology, T58.5-58.64, Electronic computers. Computer science, QA75.5-76.95
الوصف: Abstract Automated detection of defects on metal surfaces is crucial for ensuring quality control. However, the scarcity of labeled datasets for emerging target defects poses a significant obstacle. This study proposes a self-supervised representation-learning model that effectively addresses this limitation by leveraging both labeled and unlabeled data. The proposed model was developed based on a contrastive learning framework, supported by an augmentation pipeline and a lightweight convolutional encoder. The effectiveness of the proposed approach for representation learning was evaluated using an unlabeled pretraining dataset created from three benchmark datasets. Furthermore, the performance of the proposed model was validated using the NEU metal surface-defect dataset. The results revealed that the proposed method achieved a classification accuracy of 97.78%, even with fewer trainable parameters than the benchmark models. Overall, the proposed model effectively extracted meaningful representations from unlabeled image data and can be employed in downstream tasks for steel defect classification to improve quality control and reduce inspection costs.
وصف الملف: electronic resource
العلاقة: https://doaj.org/toc/2196-1115Test
-
2صورة
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: (1) Backbone uses YOLOv5s-6.0 version. (2) The Neck section adds an EVC module to replace the Transformer. (3) The Predicition module uses four detection heads.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.g001Test
https://figshare.com/articles/figure/CFM-YOLOv5_overall_structure_diagram_/24767213Test -
3
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: Aiming at the problem of low efficiency of manual detection in the field of metal surface defect detection, a deep learning defect detection method based on improved YOLOv5 algorithm is proposed. Firstly, in the feature enhancement part, we replace the multi-head self-attention module of the standard transformer encoder with the EVC module to improve the feature extraction ability. Second, in the prediction part, adding a small target detection head can reduce the negative impact of drastic object scale changes and improve the accuracy and stability of detection. Finally, the performance of the algorithm is verified by ablation experiments and analogy experiments. The experimental results show that the improved algorithm has greatly improved mAP and FPS on the data set, and can quickly and accurately identify the types of metal surface defects, which has reference significance for practical industrial applications.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.t002Test
https://figshare.com/articles/dataset/Performance_of_CFM-YOLOv5_model_/24767234Test -
4
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: Different network models for metal surface defect detection performance.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.t005Test
https://figshare.com/articles/dataset/Different_network_models_for_metal_surface_defect_detection_performance_/24767243Test -
5صورة
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: Aiming at the problem of low efficiency of manual detection in the field of metal surface defect detection, a deep learning defect detection method based on improved YOLOv5 algorithm is proposed. Firstly, in the feature enhancement part, we replace the multi-head self-attention module of the standard transformer encoder with the EVC module to improve the feature extraction ability. Second, in the prediction part, adding a small target detection head can reduce the negative impact of drastic object scale changes and improve the accuracy and stability of detection. Finally, the performance of the algorithm is verified by ablation experiments and analogy experiments. The experimental results show that the improved algorithm has greatly improved mAP and FPS on the data set, and can quickly and accurately identify the types of metal surface defects, which has reference significance for practical industrial applications.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.g002Test
https://figshare.com/articles/figure/EVC_module_structure_diagram_/24767216Test -
6صورة
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: (a) crazing, (b) inclusion, (c) patches, (d) pitted surface, (e) rolled-in scale, (f) scratches.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.g003Test
https://figshare.com/articles/figure/Six_types_of_metal_surface_defects_/24767219Test -
7صورة
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: Aiming at the problem of low efficiency of manual detection in the field of metal surface defect detection, a deep learning defect detection method based on improved YOLOv5 algorithm is proposed. Firstly, in the feature enhancement part, we replace the multi-head self-attention module of the standard transformer encoder with the EVC module to improve the feature extraction ability. Second, in the prediction part, adding a small target detection head can reduce the negative impact of drastic object scale changes and improve the accuracy and stability of detection. Finally, the performance of the algorithm is verified by ablation experiments and analogy experiments. The experimental results show that the improved algorithm has greatly improved mAP and FPS on the data set, and can quickly and accurately identify the types of metal surface defects, which has reference significance for practical industrial applications.
العلاقة: https://figshare.com/articles/figure/The_mosica_data_enhancement_schematic_diagram_/24767222Test
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.g004Test
https://figshare.com/articles/figure/The_mosica_data_enhancement_schematic_diagram_/24767222Test -
8صورة
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: The positioning loss, confidence loss and classification loss data obtained from the experimental training results.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.g005Test
https://figshare.com/articles/figure/The_positioning_loss_confidence_loss_and_classification_loss_data_obtained_from_the_experimental_training_results_/24767225Test -
9صورة
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: (a) The traditional YOLOv5 prediction box for defect types. (b) CFM-YOLOv5 algorithm for defect type prediction box.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.g006Test
https://figshare.com/articles/figure/The_prediction_results_of_deep_learning_algorithm_for_defect_data_sets_/24767228Test -
10
المؤلفون: Yuntao Xu, Peigang Jiao, Jiaqi LIU
مصطلحات موضوعية: Science Policy, Space Science, Biological Sciences not elsewhere classified, Chemical Sciences not elsewhere classified, Information Systems not elsewhere classified,
+aiming%22">xlink "> aiming, standard transformer encoder, metal surface defects, feature extraction ability, experimental results show, greatly improved map, feature enhancement part, improved yolov5 algorithm, improved algorithm, prediction part, reference significance, negative impact, low efficiency, head self, evc module, data set, cfpnet moudle, attention module, analogy experiments, accurately identify, ablation experiments الوصف: Comparison of different data enhancement experiments.
الإتاحة: https://doi.org/10.1371/journal.pone.0289179.t001Test
https://figshare.com/articles/dataset/Comparison_of_different_data_enhancement_experiments_/24767231Test
النص الكامل