Latent Diffusion Models to Enhance the Performance of Visual Defect Segmentation Networks in Steel Surface Inspection
Sensors · 18.09.2024 · 24 (18) · 6016
DOI: 10.3390/s24186016 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 18.09.2024. Verfasst von Jon Leiñena und 2 weiteren Autor:innen. Beteiligte Einrichtung: Vicomtech Foundation, Basque Research and Technology Alliance (BRTA), 20009 Donostia, Spain. Die Arbeit wurde bisher 26-mal zitiert (Crossref) und nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Elkartek Programme, Basque Government (Spain).
Abstract (Originalsprache)
This paper explores the use of state-of-the-art latent diffusion models, specifically stable diffusion, to generate synthetic images for improving the robustness of visual defect segmentation in manufacturing components. Given the scarcity and imbalance of real-world defect data, synthetic data generation offers a promising solution for training deep learning models. We fine-tuned stable diffusion using the LoRA technique on the NEU-seg dataset and evaluated the impact of different ratios of synthetic to real images on the training set of DeepLabV3+ and FPN segmentation models. Our results demonstrated a significant improvement in mean Intersection over Union (mIoU) when the training dataset was augmented with synthetic images. This study highlights the potential of diffusion models for enhancing the quality and diversity of training data in industrial defect detection, leading to more accurate and reliable segmentation results. The proposed approach achieved improvements of 5.95% and 6.85% in mIoU of defect segmentation on each model over the original dataset.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Vicomtech Foundation, Basque Research and Technology Alliance (BRTA), 20009 Donostia, Spain.
ORCID-Profile: Jon Leiñena, Fátima A Saiz
Förderung
- Elkartek Programme, Basque Government (Spain)
Zitieren
APA
Leiñena, J., Saiz, F. A., & Barandiaran, I. (2024). Latent Diffusion Models to Enhance the Performance of Visual Defect Segmentation Networks in Steel Surface Inspection. Sensors, 24(18), 6016. https://doi.org/10.3390/s24186016
Vancouver
Leiñena J, Saiz FA, Barandiaran I. Latent Diffusion Models to Enhance the Performance of Visual Defect Segmentation Networks in Steel Surface Inspection. Sensors. 2024;24(18):6016. doi:10.3390/s24186016
BibTeX
@article{Leiena2024,
author = {Leiñena, Jon and Saiz, Fátima A and Barandiaran, Iñigo},
title = {Latent Diffusion Models to Enhance the Performance of Visual Defect Segmentation Networks in Steel Surface Inspection},
journal = {Sensors},
year = {2024},
volume = {24},
number = {18},
pages = {6016},
doi = {10.3390/s24186016}
}
EUVIMED