Vision applications are becoming increasingly important for product quality surveillance in manufacturing. Training consistently well-performing visual detection algorithms based on convolutional neural networks is very challenging. Typically, there is too much training data for engineers to keep track of possible gaps in it. But even small cases of missing training data e.g. certain viewing angles can lead to trained CNNs that are unable to detect objects, that seem obvious to engineers i.e. cognition gaps. This paper presents how synthetic training data can be created in a targeted manner to close cognitive gaps of a CNN for specific use-cases. The proposed methodology uses 3D rendering to create new image data by variating scene parameters. The created data is used to reveal a cognition gap of a CNN. We show that by using this created synthetic data to train the CNN the cognition gap can be successfully closed. This is evaluated with the well-known AlexNet CNN used as a visual bicycle detector. The bicycle example is used as a stand-in for a geometrically interesting, but simple product, that is manufactured in large and growing amounts.
%0 Generic
%1 vietz2022synthetic
%A Vietz, Hannes
%A Rauch, Tristan
%A Weyrich, Michael
%B 2022 IEEE 27th International Conference on Emerging Technologies and Factory Automation (ETFA)
%D 2022
%K 2022ias ias
%T Synthetic Training Data Generation for Convolutional Neural Networks in Vision Applications
%X Vision applications are becoming increasingly important for product quality surveillance in manufacturing. Training consistently well-performing visual detection algorithms based on convolutional neural networks is very challenging. Typically, there is too much training data for engineers to keep track of possible gaps in it. But even small cases of missing training data e.g. certain viewing angles can lead to trained CNNs that are unable to detect objects, that seem obvious to engineers i.e. cognition gaps. This paper presents how synthetic training data can be created in a targeted manner to close cognitive gaps of a CNN for specific use-cases. The proposed methodology uses 3D rendering to create new image data by variating scene parameters. The created data is used to reveal a cognition gap of a CNN. We show that by using this created synthetic data to train the CNN the cognition gap can be successfully closed. This is evaluated with the well-known AlexNet CNN used as a visual bicycle detector. The bicycle example is used as a stand-in for a geometrically interesting, but simple product, that is manufactured in large and growing amounts.
@conference{vietz2022synthetic,
abstract = {Vision applications are becoming increasingly important for product quality surveillance in manufacturing. Training consistently well-performing visual detection algorithms based on convolutional neural networks is very challenging. Typically, there is too much training data for engineers to keep track of possible gaps in it. But even small cases of missing training data e.g. certain viewing angles can lead to trained CNNs that are unable to detect objects, that seem obvious to engineers i.e. cognition gaps. This paper presents how synthetic training data can be created in a targeted manner to close cognitive gaps of a CNN for specific use-cases. The proposed methodology uses 3D rendering to create new image data by variating scene parameters. The created data is used to reveal a cognition gap of a CNN. We show that by using this created synthetic data to train the CNN the cognition gap can be successfully closed. This is evaluated with the well-known AlexNet CNN used as a visual bicycle detector. The bicycle example is used as a stand-in for a geometrically interesting, but simple product, that is manufactured in large and growing amounts.},
added-at = {2022-10-31T11:14:51.000+0100},
author = {Vietz, Hannes and Rauch, Tristan and Weyrich, Michael},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2f5f350e2dba6a3ad823d7cffc232c66e/taylansngerli},
booktitle = {2022 IEEE 27th International Conference on Emerging Technologies and Factory Automation (ETFA)},
interhash = {ef9361fbe13c34c191e655cc6fbd766c},
intrahash = {f5f350e2dba6a3ad823d7cffc232c66e},
keywords = {2022ias ias},
timestamp = {2022-10-31T10:14:51.000+0100},
title = {Synthetic Training Data Generation for Convolutional Neural Networks in Vision Applications},
year = 2022
}
