An End-To-End Bayesian Segmentation Network Based on a Generative Adversarial Network for Remote Sensing Images

Xinlong Liu,Chu He,Dehui Xiong,Mingsheng Liao

doi:10.3390/rs12020216

Abstract

Due to the development of deep convolutional neural networks (CNNs), great progress has been made in semantic segmentation recently. In this paper, we present an end-to-end Bayesian segmentation network based on generative adversarial networks (GANs) for remote sensing images. First, fully convolutional networks (FCNs) and GANs are utilized to realize the derivation of the prior probability and the likelihood to the posterior probability in Bayesian theory. Second, the cross-entropy loss in the FCN serves as an a priori to guide the training of the GAN, so as to avoid the problem of mode collapse during the training process. Third, the generator of the GAN is used as a teachable spatial filter to construct the spatial relationship between each label. Some experiments were performed on two remote sensing datasets, and the results demonstrate that the training of the proposed method is more stable than other GAN based models. The average accuracy and mean intersection (MIoU) of the two datasets were 0.0465 and 0.0821, and 0.0772 and 0.1708 higher than FCN, respectively.

Highlights

In order to verify the stability of the proposed method in training, we compared it with other generative adversarial networks (GANs)-based segmentation methods, such as pix2pix [30] and CRFAS [35]
The second comparison algorithm was Deeplab, since it essentially optimizes the results of the fully convolutional networks (FCNs) using conditional random field (CRF), which is similar to the proposed approach
This paper has presented an end-to-end Bayesian segmentation network based on a generative adversarial network for remote sensing images

Summary

Introduction

Many scene recognition applications, such as virtual or augmented reality, human–computer intersections [1], and autonomous driving [2,3], urgently need accurate and effective segmentation mechanisms Driven by the these requirements, image semantic segmentation has gained more and more attention from machine learning and computer vision researchers. In 2014, Long et al proposed fully convolutional network (FCN) [10] which popularized the architectures of CNN for semantic segmentation without fully connected layers. This paradigm was adopted by almost all subsequent state of the art segmentation approaches. Many classic networks, such as AlexNet [11], VGG-16 [12], GoogLeNet [13], and ResNet [14] have been used as basic modules for many semantic segmentation architectures

Methods

Results

Discussion

Conclusion