Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders

Paul Bergmann, Sindy Löwe, Michael Fauser, David Sattlegger, Carsten Steger

2019

Abstract

Convolutional autoencoders have emerged as popular methods for unsupervised defect segmentation on image data. Most commonly, this task is performed by thresholding a per-pixel reconstruction error based on an p̀-distance. This procedure, however, leads to large residuals whenever the reconstruction includes slight localization inaccuracies around edges. It also fails to reveal defective regions that have been visually altered when intensity values stay roughly consistent. We show that these problems prevent these approaches from being applied to complex real-world scenarios and that they cannot be easily avoided by employing more elaborate architectures such as variational or feature matching autoencoders. We propose to use a perceptual loss function based on structural similarity that examines inter-dependencies between local image regions, taking into account luminance, contrast, and structural information, instead of simply comparing single pixel values. It achieves significant performance gains on a challenging real-world dataset of nanofibrous materials and a novel dataset of two woven fabrics over state-of-the-art approaches for unsupervised defect segmentation that use per-pixel reconstruction error metrics.

Paper Citation

in Harvard Style

Bergmann P., Löwe S., Fauser M., Sattlegger D. and Steger C. (2019). Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders. In Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2019) - Volume 5: VISAPP; ISBN 978-989-758-354-4, SciTePress, pages 372-380. DOI: 10.5220/0007364503720380

in Bibtex Style

@conference{visapp19,
author={Paul Bergmann and Sindy Löwe and Michael Fauser and David Sattlegger and Carsten Steger},
title={Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders},
booktitle={Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2019) - Volume 5: VISAPP},
year={2019},
pages={372-380},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0007364503720380},
isbn={978-989-758-354-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2019) - Volume 5: VISAPP
TI - Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders
SN - 978-989-758-354-4
AU - Bergmann P.
AU - Löwe S.
AU - Fauser M.
AU - Sattlegger D.
AU - Steger C.
PY - 2019
SP - 372
EP - 380
DO - 10.5220/0007364503720380
PB - SciTePress