Deep Structured Features for Semantic Segmentation

Michael Tschannen, Lukas Cavigelli, Fabian Mentzer, Thomas Wiatowski, Luca Benini

Published 2016-09-26Version 1

We propose a highly structured neural network architecture for semantic segmentation of images that combines i) a Haar wavelet-based tree-like convolutional neural network (CNN), ii) a random layer realizing a radial basis function kernel approximation, and iii) a linear classifier. While stages i) and ii) are completely pre-specified, only the linear classifier is learned from data. Thanks to its high degree of structure, our architecture has a very small memory footprint and thus fits onto low-power embedded and mobile platforms. We apply the proposed architecture to outdoor scene and aerial image semantic segmentation and show that the accuracy of our architecture is competitive with conventional pixel classification CNNs. Furthermore, we demonstrate that the proposed architecture is data efficient in the sense of matching the accuracy of pixel classification CNNs when trained on a much smaller data set.