Advanced search
1 file | 5.29 MB Add to list

Deep learning for paint loss detection with a multiscale, translation invariant network

Author
Organization
Abstract
We explore the potential of deep learning in digital painting analysis to facilitate condition reporting and to support restoration treatments. We address the problem of paint loss detection and develop a multiscale deep learning system with dilated convolutions that enables a large receptive field with limited training parameters to avoid overtraining. Our model handles efficiently multimodal data that are typically acquired in art investigation. As a case study we use multimodal data of the Ghent Altarpiece. Our results indicate huge potential of the proposed approach in terms of accuracy and also its fast execution, which allows interactivity and continuous learning.
Keywords
Art investigation, paint loss, multi-modal data, semantic segmentation, deep learning, transfer learning

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 5.29 MB

Citation

Please use this url to cite or link to this publication:

MLA
Meeus, Laurens et al. “Deep Learning for Paint Loss Detection with a Multiscale, Translation Invariant Network.” Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019). 2019. Print.
APA
Meeus, L., Huang, S., Devolder, B., Dubois, H., & Pizurica, A. (2019). Deep learning for paint loss detection with a multiscale, translation invariant network. Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019). Presented at the 11th Int’l Symposium on Image and Signal Processing and Analysis (ISPA 2019).
Chicago author-date
Meeus, Laurens, Shaoguang Huang, Bart Devolder, Hélène Dubois, and Aleksandra Pizurica. 2019. “Deep Learning for Paint Loss Detection with a Multiscale, Translation Invariant Network.” In Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019).
Chicago author-date (all authors)
Meeus, Laurens, Shaoguang Huang, Bart Devolder, Hélène Dubois, and Aleksandra Pizurica. 2019. “Deep Learning for Paint Loss Detection with a Multiscale, Translation Invariant Network.” In Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019).
Vancouver
1.
Meeus L, Huang S, Devolder B, Dubois H, Pizurica A. Deep learning for paint loss detection with a multiscale, translation invariant network. Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019). 2019.
IEEE
[1]
L. Meeus, S. Huang, B. Devolder, H. Dubois, and A. Pizurica, “Deep learning for paint loss detection with a multiscale, translation invariant network,” in Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019), Dubrovnik, Croatia, 2019.
@inproceedings{8624037,
  abstract     = {We explore the potential of deep learning in digital painting analysis to facilitate condition reporting and to support restoration treatments. We address the problem of paint loss detection and develop a multiscale deep learning system with dilated convolutions that enables a large receptive field with limited training parameters to avoid overtraining. Our model handles efficiently multimodal data that are typically acquired in art investigation. As a case study we use multimodal data of the Ghent Altarpiece. Our results indicate huge potential of the proposed approach in terms of accuracy and also its fast execution, which allows interactivity and continuous learning.},
  author       = {Meeus, Laurens and Huang, Shaoguang and Devolder, Bart and Dubois, Hélène and Pizurica, Aleksandra},
  booktitle    = {Proceedings of the 11th International Symposium on Image and Signal Processing and Analysis (ISPA 2019)},
  keywords     = {Art investigation,paint loss,multi-modal data,semantic segmentation,deep learning,transfer learning},
  language     = {eng},
  location     = {Dubrovnik, Croatia},
  pages        = {5},
  title        = {Deep learning for paint loss detection with a multiscale, translation invariant network},
  year         = {2019},
}