Rotation invariant filters in CNNs

applied to segmentation of aerial images for land-use classification

Master Thesis (2018)
Author(s)

A. Dhar (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Contributor(s)

Jan van Van Gemert – Mentor

Marcel JT Reinders – Graduation committee member

Odette Scharenborg – Graduation committee member

Daan van der Maas – Mentor

Faculty
Electrical Engineering, Mathematics and Computer Science
Copyright
© 2018 Aniket Dhar
More Info
expand_more
Publication Year
2018
Language
English
Copyright
© 2018 Aniket Dhar
Graduation Date
21-11-2018
Awarding Institution
Delft University of Technology
Programme
['Electrical Engineering | Embedded Systems']
Faculty
Electrical Engineering, Mathematics and Computer Science
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Convolutional neural networks are showing incredible performance in image classification, segmentation, object detection and other computer vision applications in recent years. But they lack understanding of affine transformations to input data. In this work, we introduce rotational invariant
convolutional neural networks that learn rotational invariance by design, and not from data. We build rotation invariant filters through parametric learning of linear combination of a basis set of filters, rather than modelling the filters ourselves. Our approach combines the learning capability of CNNs with custom filter selection. We show stability in performance under rotations in input images. We first validate our findings for classification on MNIST and then for
multi-class semantic segmentation on the DeepGlobe 2018 Satellite Image Understanding Challenge.

Files

License info not available