Application of Cellular Neural Networks in Semantic Segmentation

Abstract

The popularity of convolutional neural networks and deep learning based approaches has increased continuously in the past years. These methods has enabled the solution of various practical problems, but they still are not heavily exploited in the embedded domain, which requires low-power implementation of these architectures. Cellular neural networks can provide an analogue and power-efficient implementation of these networks which also enables the exploitation of non-Boolean, beyond CMOS elements such as memristive structures.In this paper we will demonstrate on simple datasets how cellular neural networks with different cell dynamics can be efficiently applied in the problem of semantic segmentation.

The deep multichannel discrete‐time cellular neural network model for classification

“…Fulop et al demonstrated that CellNNs can be used for semantic segmentation 2 and image classification 22 as well. Classification accuracy of their networks is not included in Table 5, since the focus of their paper is obtaining templates for memristive continuous‐time CellNNs, while this paper is about discrete‐time CellNNs.…”

“…Studies on deep learning have grown exponentially in the last decade. Some of the application fields that benefit from deep learning are speech recognition, 1 semantic segmentation, 2 visual object recognition, 3 and many other domains such as drug discovery 4 and genomics 5 . Krizhevsky et al 6 have made a major breakthrough in computer vision by using convolutional neural networks (CNNs) by introducing AlexNET.…”