A Survey on Deep Learning-based Architectures for Semantic Segmentation on 2D Images

Abstract

Semantic segmentation is the pixel-wise labelling of an image. Since the problem is defined at the pixel level, determining image class labels only is not acceptable, but localising them at the original image pixel resolution is necessary. Boosted by the extraordinary ability of convolutional neural networks (CNN) in creating semantic, high level and hierarchical image features; excessive numbers of deep learning-based 2D semantic segmentation approaches have been proposed within the last decade. In this survey, we mainly focus on the recent scientific developments in semantic segmentation, specifically on deep learning-based methods using 2D images. We started with an analysis of the public image sets and leaderboards for 2D semantic segmantation, with an overview of the techniques employed in performance evaluation. In examining the evolution of the field, we chronologically categorised the approaches into three main periods, namely pre-and early deep learning era, the fully convolutional era, and the post-FCN era. We technically analysed the solutions put forward in terms of solving the fundamental problems of the field, such as fine-grained localisation and scale invariance. Before drawing our conclusions, we present a table of methods from all mentioned eras, with a brief summary of each approach that explains their contribution to the field. We conclude the survey by discussing the current challenges of the field and to what extent they have been solved.

Shape-Constrained Method of Remote Sensing Monitoring of Marine Raft Aquaculture Areas on Multitemporal Synthetic Sentinel-1 Imagery

“…The refined extraction task of the marine raft aquaculture area aims to identify all of the pixels belonging to marine raft aquaculture areas in the image and locate their positions to serve the subsequent statistics of aquaculture area and estimation of aquaculture production. This task belongs to the full-pixel semantic segmentation problem in the field of computer vision [29][30][31][32]. As popular semantic segmentation models, FCN [33], Unet [34], PSPnet [35] and DeepLabV3+ [36] each have their own advantages.…”

A Spacecraft Dataset for Detection, Segmentation and Parts Recognition

“…The training batch sizes for DeepLabV3+ Xception were 8 while the rest of models had batch size 4. We use pixel accuracy (PixAcc) [39] and mean intersection-of-union (mIoU) [40] to compare model performances.…”

FuSS: Fusing Superpixels for Improved Segmentation Consistency

“…Many different methods have been developed to solve this problem: FCN [6], U-Net [7], SegNet [8], among others. In recent years, many models have significantly improved the quality of semantic segmentation [9], [10].…”