Next Article in Journal
Automated Detection of Animals in Low-Resolution Airborne Thermal Imagery
Next Article in Special Issue
AAU-Net: Attention-Based Asymmetric U-Net for Subject-Sensitive Hashing of Remote Sensing Images
Previous Article in Journal
The MSG Technique: Improving Commercial Microwave Link Rainfall Intensity by Using Rain Area Detection from Meteosat Second Generation
Previous Article in Special Issue
Automatic Extraction of Water Inundation Areas Using Sentinel-1 Data for Large Plain Areas
 
 
Technical Note

MARE: Self-Supervised Multi-Attention REsu-Net for Semantic Segmentation in Remote Sensing

1
Department of Computer, Control and Management Engineering Antonio Ruberti (DIAG), Sapienza University of Rome, 00185 Rome, Italy
2
Department of Information Engineering, Electronics and Telecommunication, Sapienza University of Rome, 00184 Rome, Italy
3
Computer Science Department, University of Crete, 70013 Heraklion, Greece
*
Author to whom correspondence should be addressed.
Academic Editors: Amir Hussain, Ahmed Al-Dubai, William (Bill) J Buchanan, Jonathan Wu, Kaizhu Huang, Bin Luo, Jin Tang, Wadii Boulila and Adel M. Alimi
Remote Sens. 2021, 13(16), 3275; https://doi.org/10.3390/rs13163275
Received: 5 July 2021 / Revised: 6 August 2021 / Accepted: 14 August 2021 / Published: 19 August 2021
(This article belongs to the Special Issue Big Data Analytics for Secure and Smart Environmental Services)
Scene understanding of satellite and aerial images is a pivotal task in various remote sensing (RS) practices, such as land cover and urban development monitoring. In recent years, neural networks have become a de-facto standard in many of these applications. However, semantic segmentation still remains a challenging task. With respect to other computer vision (CV) areas, in RS large labeled datasets are not very often available, due to their large cost and to the required manpower. On the other hand, self-supervised learning (SSL) is earning more and more interest in CV, reaching state-of-the-art in several tasks. In spite of this, most SSL models, pretrained on huge datasets like ImageNet, do not perform particularly well on RS data. For this reason, we propose a combination of a SSL algorithm (particularly, Online Bag of Words) and a semantic segmentation algorithm, shaped for aerial images (namely, Multistage Attention ResU-Net), to show new encouraging results (i.e., 81.76% mIoU with ResNet-18 backbone) on the ISPRS Vaihingen dataset. View Full-Text
Keywords: semantic segmentation; self-supervised learning; linear attention; Vaihingen dataset semantic segmentation; self-supervised learning; linear attention; Vaihingen dataset
Show Figures

Graphical abstract

MDPI and ACS Style

Marsocci, V.; Scardapane, S.; Komodakis, N. MARE: Self-Supervised Multi-Attention REsu-Net for Semantic Segmentation in Remote Sensing. Remote Sens. 2021, 13, 3275. https://doi.org/10.3390/rs13163275

AMA Style

Marsocci V, Scardapane S, Komodakis N. MARE: Self-Supervised Multi-Attention REsu-Net for Semantic Segmentation in Remote Sensing. Remote Sensing. 2021; 13(16):3275. https://doi.org/10.3390/rs13163275

Chicago/Turabian Style

Marsocci, Valerio, Simone Scardapane, and Nikos Komodakis. 2021. "MARE: Self-Supervised Multi-Attention REsu-Net for Semantic Segmentation in Remote Sensing" Remote Sensing 13, no. 16: 3275. https://doi.org/10.3390/rs13163275

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop