Deep Structured Features for Semantic Segmentation

TL;DR

This paper introduces a compact neural network architecture for semantic segmentation that combines wavelet-based CNNs, kernel approximation, and a linear classifier, achieving competitive accuracy with less data and smaller models.

Contribution

The paper presents a novel, highly structured neural network architecture that is highly efficient and suitable for low-power devices, with a unique combination of pre-specified layers and a learned classifier.

Findings

Achieves competitive accuracy with traditional CNNs on outdoor and aerial images.

Requires significantly less training data to reach similar performance.

Has a very small model size suitable for embedded platforms.

Abstract

We propose a highly structured neural network architecture for semantic segmentation with an extremely small model size, suitable for low-power embedded and mobile platforms. Specifically, our architecture 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. 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.