Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

A Step towards Explainable Artificial Neural Networks in Image Processing by Dataset Assessment

: Heide, Nina Felicitas; Albrecht, Alexander; Heizmann, Michael

Fulltext urn:nbn:de:0011-n-6088552 (199 KByte PDF)
MD5 Fingerprint: 6ae5a6c057d85713815faf362e01d764
Created on: 21.11.2020

Heizmann, Michael (Hrsg.); Längle, Thomas (Hrsg.):
Forum Bildverarbeitung 2020 : 26. und 27. November 2020, Karlsruhe, Online-Konferenz
Karlsruhe: KIT Scientific Publishing, 2020
ISBN: 978-3-7315-1053-6
DOI: 10.5445/KSP/1000124383
Forum Bildverarbeitung <2020, Online>
Conference Paper, Electronic Publication
Fraunhofer IOSB ()
artificial neural network; image processing; premodeling explainability; robot vision system

We propose a methodology for generalized exploratory data analysis focusing on artificial neural network (ANN) methods. Our method is denoted IC-ACC due to the combined assessment of information content (IC) and accuracy (ACC) and aims at answering a frequently posed question in ANN research: ”What is good data?” As the dataset has the primary influence on the development of the model, IC-ACC provides a step towards explainable ANN methods in the pre-modeling stage by a better insight in the dataset. With this insight, detrimental data can be eliminated before a negative influence on the ANN performance occurs. IC-ACC constitutes a guideline to generate efficient and accurate data for a specific, data-driven ANN method. Moreover, we show that training an ANN for the semantic segmentation of 3Ddata from unstructured environments with IC-ACC-assessed and -customized training data contributes to a more efficient training. The IC-ACC method is demonstrated on application examples for the visual perception of robotic platforms.