Summary of Diffusion-based Image Generation For In-distribution Data Augmentation in Surface Defect Detection, by Luigi Capogrosso et al.
Diffusion-based Image Generation for In-distribution Data Augmentation in Surface Defect Detection
by Luigi Capogrosso, Federico Girella, Francesco Taioli, Michele Dalla Chiara, Muhammad Aqeel, Franco Fummi, Francesco Setti, Marco Cristani
First submitted to arxiv on: 1 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This study demonstrates the effectiveness of diffusion models in improving data augmentation for surface defect detection. By leveraging diffusion models, researchers can generate more realistic in-distribution defects, enabling classification systems to learn genuine defect appearances. The proposed In&Out approach combines out-of-distribution and in-distribution samples for data augmentation, addressing zero-shot, few-shot, and full-shot scenarios. Experimental results on the Kolektor Surface-Defect Dataset 2 show a new state-of-the-art classification AP score of .782 under weak supervision. This paper contributes to the development of more robust defect detection classifiers by providing a novel approach for data augmentation using diffusion models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This study shows how to make computers better at detecting defects on surfaces. Currently, computers are trained using real and fake images, but this can be limited. The researchers found that using “diffusion models” helps create more realistic fake images, which makes the computer learn what a real defect looks like. They proposed a new way of creating these fake images, called In&Out, which works in different situations. The results show that their approach is better than previous methods, and it can be used to improve defect detection. |
Keywords
» Artificial intelligence » Classification » Data augmentation » Diffusion » Few shot » Zero shot
