FSAU-Net: a network for extracting buildings from remote sensing imagery using feature self-attention

Convolutional neural networks (CNNs) extract semantic features from images by stacking convolutional operators, which easily causes semantic information loss and leads to hollow and edge inaccuracies in building extraction. Therefore, a features self-attention U-block network (FSAU-Net) is proposed....

Full description

Bibliographic Details

Published in:	International journal of remote sensing Vol. 44; no. 5; pp. 1643 - 1664
Main Authors:	Hu, Minghong, Li, Jiatian, A, Xiaohui, Zhao, Yunfei, Lu, Mei, Li, Wen
Format:	Journal Article
Language:	English
Published:	London Taylor & Francis 04-03-2023 Taylor & Francis Ltd
Subjects:	Ablation Artificial neural networks Building extraction Buildings Coders Features self-attention Modules Neural networks Remote sensing Resolution Semantics Skip connection Spatial attention
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!