A beneficial dual transformation approach for deep learning networks used in steel surface defect detection

Fityanul Akhyar, Chih Yang Lin, Gugan S. Kathiresan

研究成果: 書貢獻/報告類型會議論文篇章同行評審

7 引文 斯高帕斯(Scopus)

摘要

Steel surface defect detection represents a challenging task in real-world practical object detection. Based on our observations, there are two critical problems which create this challenge: the tiny size, and vagueness of the defects. To solve these problems, this study a proposes a deep learning-based defect detection system that uses automatic dual transformation in the end-to-end network. First, the original training images in RGB are transformed into the HSV color model to re-arrange the difference in color distribution. Second, the feature maps are upsampled using bilinear interpolation to maintain the smaller resolution. The latest and state-of-the-art object detection model, High-Resolution Network (HRNet) is utilized in this system, with initial transformation performed via data augmentation. Afterward, the output of the backbone stage is applied to the second transformation. According to the experimental results, the proposed approach increases the accuracy of the detection of class 1 Severstal steel surface defects by 3.6% versus the baseline.

原文???core.languages.en_GB???
主出版物標題ICMR 2021 - Proceedings of the 2021 International Conference on Multimedia Retrieval
發行者Association for Computing Machinery, Inc
頁面619-622
頁數4
ISBN(電子)9781450384636
DOIs
出版狀態已出版 - 24 8月 2021
事件11th ACM International Conference on Multimedia Retrieval, ICMR 2021 - Taipei, Taiwan
持續時間: 16 11月 202119 11月 2021

出版系列

名字ICMR 2021 - Proceedings of the 2021 International Conference on Multimedia Retrieval

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???11th ACM International Conference on Multimedia Retrieval, ICMR 2021
國家/地區Taiwan
城市Taipei
期間16/11/2119/11/21

指紋

深入研究「A beneficial dual transformation approach for deep learning networks used in steel surface defect detection」主題。共同形成了獨特的指紋。

引用此