The detection of manufacturing errors is crucial in fabrication processes to
ensure product quality and safety standards. Since many defects occur very
rarely and their characteristics are mostly unknown a priori, their detection
is still an open research question. To this end, we propose DifferNet: It
leverages the descriptiveness of features extracted by convolutional neural
networks to estimate their density using normalizing flows. Normalizing flows
are well-suited to deal with low dimensional data distributions. However, they
struggle with the high dimensionality of images. Therefore, we employ a
multi-scale feature extractor which enables the normalizing flow to assign
meaningful likelihoods to the images. Based on these likelihoods we develop a
scoring function that indicates defects. Moreover, propagating the score back
to the image enables pixel-wise localization. To achieve a high robustness and
performance we exploit multiple transformations in training and evaluation. In
contrast to most other methods, ours does not require a large number of
training samples and performs well with as low as 16 images. We demonstrate the
superior performance over existing approaches on the challenging and newly
proposed MVTec AD and Magnetic Tile Defects datasets.
Description
[2008.12577] Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flows
%0 Generic
%1 rudolph2020differnet
%A Rudolph, Marco
%A Wandt, Bastian
%A Rosenhahn, Bodo
%D 2020
%K l3s leibnizailab
%T Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flows.
%U http://arxiv.org/abs/2008.12577
%X The detection of manufacturing errors is crucial in fabrication processes to
ensure product quality and safety standards. Since many defects occur very
rarely and their characteristics are mostly unknown a priori, their detection
is still an open research question. To this end, we propose DifferNet: It
leverages the descriptiveness of features extracted by convolutional neural
networks to estimate their density using normalizing flows. Normalizing flows
are well-suited to deal with low dimensional data distributions. However, they
struggle with the high dimensionality of images. Therefore, we employ a
multi-scale feature extractor which enables the normalizing flow to assign
meaningful likelihoods to the images. Based on these likelihoods we develop a
scoring function that indicates defects. Moreover, propagating the score back
to the image enables pixel-wise localization. To achieve a high robustness and
performance we exploit multiple transformations in training and evaluation. In
contrast to most other methods, ours does not require a large number of
training samples and performs well with as low as 16 images. We demonstrate the
superior performance over existing approaches on the challenging and newly
proposed MVTec AD and Magnetic Tile Defects datasets.
@misc{rudolph2020differnet,
abstract = {The detection of manufacturing errors is crucial in fabrication processes to
ensure product quality and safety standards. Since many defects occur very
rarely and their characteristics are mostly unknown a priori, their detection
is still an open research question. To this end, we propose DifferNet: It
leverages the descriptiveness of features extracted by convolutional neural
networks to estimate their density using normalizing flows. Normalizing flows
are well-suited to deal with low dimensional data distributions. However, they
struggle with the high dimensionality of images. Therefore, we employ a
multi-scale feature extractor which enables the normalizing flow to assign
meaningful likelihoods to the images. Based on these likelihoods we develop a
scoring function that indicates defects. Moreover, propagating the score back
to the image enables pixel-wise localization. To achieve a high robustness and
performance we exploit multiple transformations in training and evaluation. In
contrast to most other methods, ours does not require a large number of
training samples and performs well with as low as 16 images. We demonstrate the
superior performance over existing approaches on the challenging and newly
proposed MVTec AD and Magnetic Tile Defects datasets.},
added-at = {2021-07-19T15:11:44.000+0200},
author = {Rudolph, Marco and Wandt, Bastian and Rosenhahn, Bodo},
biburl = {https://www.bibsonomy.org/bibtex/2b85788dc12bd57f519305859b7e2cf7d/sophieschr},
description = {[2008.12577] Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flows},
interhash = {d42ba926baa739c1d6f85fdea13634b8},
intrahash = {b85788dc12bd57f519305859b7e2cf7d},
keywords = {l3s leibnizailab},
month = aug,
note = {cite arxiv:2008.12577},
timestamp = {2021-07-19T15:11:44.000+0200},
title = {Same Same But DifferNet: Semi-Supervised Defect Detection with Normalizing Flows.},
url = {http://arxiv.org/abs/2008.12577},
year = 2020
}