A multi-stage active learning framework with an instance-based sample selection algorithm for steel surface defect
Authors:
- Shuo GaoSGShuo Gao
- Yimin JiangYJYimin Jiang
- Tangbin XiaTXTangbin Xia
- Yaping LiYLYaping Li
- Ying ZhuYZYing Zhu
- Lifeng XiLXLifeng Xi
- 28 December 2024
- journal article
- research article
- Published by Elsevier in Advanced Engineering Informatics
- Vol. 64
Abstract
No abstract availableKeywords
Funding Information
- China Postdoctoral Science Foundation (2024T170553)
- National Key Research and Development Program of China (2022YFF0605700)
- National Natural Science Foundation of China (72301169)
This publication has 21 references indexed in Scilit:
- Prototype-guided domain adaptive one-stage object detector for defect detectionAdvanced Engineering Informatics, 2024
- A quantitative causal analysis and optimization framework for inclusions of steel productsAdvanced Engineering Informatics, 2024
- An inspection network with dynamic feature extractor and task alignment head for steel surface defectMeasurement, 2023
- A high-accuracy and lightweight detector based on a graph convolution network for strip surface defect detectionAdvanced Engineering Informatics, 2023
- Surface defect detection and classification of steel using an efficient Swin TransformerAdvanced Engineering Informatics, 2023
- Selecting informative data for defect segmentation from imbalanced datasets via active learningAdvanced Engineering Informatics, 2023
- Active learning and novel model calibration measurements for automated visual inspection in manufacturingJournal of Intelligent Manufacturing, 2023
- Deep active learning for object detectionInformation Sciences, 2021
- HAL: Hybrid active learning for efficient labeling in medical domainNeurocomputing, 2021
- Combining clustering and active learning for the detection and learning of new image classesNeurocomputing, 2019