Gans in Industrial Surface Defect Detection : Insights and Challenges
Defect detection is an essential task in industrial companies to ensure that their products meet quality and safety standards. Therefore, there is increasing interest in applying computer vision and deep learning techniques to automate the defect detection process. This study discusses the importance of defect detection in industrial companies and how computer vision based on deep learning models can be used to perform this task. Despite numerous proposed methods for defect detection, the issue of data availability and imbalanced data remains a common and natural problem in this domain. These problems can pose significant challenges to optimizing models for defect detection. Therefore, this study aims to survey all GAN-based models in industrial defect detection and explore how GANs can be implemented in this specific domain. Our work mainly focuses on industrial defect detection and aims to provide a brief and initial starting point for researchers who wish to explore this topic. We begin by formulating the main problems related to data collection, then discuss the state-of-the-art of GANs before analysing the existing solutions proposed by GAN-based models and the challenges of implementing GANs in surface defect detection. Furthermore, we present the future direction of GANs in defect detection and other related fields. The findings of this study provide a high level of information about GAN-based models in defect detection, contributing to the body of knowledge in this field
Year of publication: |
[2023]
|
---|---|
Authors: | Farady, Isack ; Islam, Junaidul ; Tuarob, Suppawong ; Ng, Hui-Fuang ; Lin, Chih-Yang |
Publisher: |
[S.l.] : SSRN |
Saved in:
freely available
Saved in favorites
Similar items by person
-
A qualitative study of obesity perceptions in social media through Twitter's tweets
Chen, I-Chi, (2018)
-
Suppawong Tuarob, (2022)
-
A qualitative study of obesity perceptions in social media through Twitter's tweets
Chen, I-Chi, (2018)
- More ...