Special Issue "Deep Learning for Remote Sensing Data"

A special issue of Remote Sensing (ISSN 2072-4292). This special issue belongs to the section "Environmental Remote Sensing".

Deadline for manuscript submissions: 25 June 2021.

Special Issue Editors

Prof. Ajmal Mian
Website
Guest Editor
The University of Western Australia
Interests: deep learning; remote sensing; hyperspectral image analysis; classification; tracking; data fusion; video analysis; 3D point cloud analysis; LiDAR data analysis
Dr. Jun Zhou
Website
Guest Editor
School of Information and Communication Technology, Griffith University, Nathan, QLD 4111, Australia
Interests: hyperspectral image preprocessing; hyperspectral image visualization; spectral–spatial feature extraction; hyperspectral unmixing; hyperspectral object detection and classification; 3D modeling and depth estimation; multimodal image fusion; applications in environmental monitoring; agriculture and medicine
Dr. Naveed Akhtar
Website
Guest Editor
University of Western Australia
Interests: deep learning; remote sensing; hyperspectral image analysis; adversarial attacks and defenses
Dr. Pedram Ghamisi
Website
Guest Editor
Prof. Antonio Robles-Kelly
Website
Guest Editor
Faculty of Sci Eng & Built Env, School of Info Technology, Geelong Waurn Ponds Campus, Deakin University, Geelong, Australia
Interests: reflectance models, pattern recognition, machine learning, computer vision, segmentation, graph-matching, imaging spectroscopy, shape-from-X, environmental management
Prof. Dr. Tat-Jun Chin
Website
Guest Editor
School of Computer Science, The University of Adelaide, Adelaide, Australia
Interests: 3D sensing, processing and analysis, LiDAR data analysis, augmented reality, large scale optimization

Special Issue Information

The past decade has seen a quantum leap in the accuracies of numerous signal and image processing tasks due to deep learning. Deep learning can model very complex nonlinear mathematical functions in a data-driven manner, which makes it an attractive technology for numerous tasks in the field of remote sensing. Moreover, the recent rise in the number of Earth-observing satellites has also resulted in large volumes of data, which makes the application of deep learning even more appealing for remote sensing data. The ever-increasing computational capacity of GPUs and efficient implementation of deep learning algorithms in public software libraries are additional factors that are currently shifting the focus of the remote sensing community towards deep learning as the main data analysis tool.

This Special Issue on “Deep Learning for Remote Sensing Data” aims to capture recent advances and trends in exploiting deep learning for complex remote sensing data analysis tasks. The Special Issue welcomes contributions towards both theoretical advancements of the deep learning framework in the context of remote sensing, as well as application of this technology to remote sensing data. The topics of interest include but are not limited to:

  • Deep learning for remote sensing image processing, e.g., pan-sharpening, super-resolution;
  • Remote sensing data analysis with deep learning;
  • Specialized network architectures and deep learning algorithms for remote sensing data;
  • Transfer learning and cross-domain learning;
  • Real and synthetic remote sensing data generation;
  • Multimodality data fusion with deep models;
  • Pixel-level and subpixel-level classification, e.g., hyperspectral unmixing, segmentation.

Keywords

  • remote sensing
  • deep learning
  • hyperspectral imaging
  • segmentation
  • pan-sharpening
  • hyperspectral unmixing

Published Papers (2 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

Jump to: Other

Open AccessArticle
High-Rankness Regularized Semi-Supervised Deep Metric Learning for Remote Sensing Imagery
Remote Sens. 2020, 12(16), 2603; https://doi.org/10.3390/rs12162603 - 12 Aug 2020
Abstract
Deep metric learning has recently received special attention in the field of remote sensing (RS) scene characterization, owing to its prominent capabilities for modeling distances among RS images based on their semantic information. Most of the existing deep metric learning methods exploit pairwise [...] Read more.
Deep metric learning has recently received special attention in the field of remote sensing (RS) scene characterization, owing to its prominent capabilities for modeling distances among RS images based on their semantic information. Most of the existing deep metric learning methods exploit pairwise and triplet losses to learn the feature embeddings with the preservation of semantic-similarity, which requires the construction of image pairs and triplets based on the supervised information (e.g., class labels). However, generating such semantic annotations becomes a completely unaffordable task in large-scale RS archives, which may eventually constrain the availability of sufficient training data for this kind of models. To address this issue, we reformulate the deep metric learning scheme in a semi-supervised manner to effectively characterize RS scenes. Specifically, we aim at learning metric spaces by utilizing the supervised information from a small number of labeled RS images and exploring the potential decision boundaries for massive sets of unlabeled aerial scenes. In order to reach this goal, a joint loss function, composed of a normalized softmax loss with margin and a high-rankness regularization term, is proposed, as well as its corresponding optimization algorithm. The conducted experiments (including different state-of-the-art methods and two benchmark RS archives) validate the effectiveness of the proposed approach for RS image classification, clustering and retrieval tasks. The codes of this paper are publicly available. Full article
(This article belongs to the Special Issue Deep Learning for Remote Sensing Data)
Show Figures

Figure 1

Other

Jump to: Research

Open AccessTechnical Note
PlaNet: A Neural Network for Detecting Transverse Aeolian Ridges on Mars
Remote Sens. 2020, 12(21), 3607; https://doi.org/10.3390/rs12213607 - 03 Nov 2020
Abstract
Transverse aeolian ridges (TARs) are unusual bedforms on the surface of Mars. TARs are common but sparse on Mars; TAR fields are small, rarely continuous, and scattered, making manual mapping impractical. There have been many efforts to automatically classify the Martian surface, but [...] Read more.
Transverse aeolian ridges (TARs) are unusual bedforms on the surface of Mars. TARs are common but sparse on Mars; TAR fields are small, rarely continuous, and scattered, making manual mapping impractical. There have been many efforts to automatically classify the Martian surface, but they have never explicitly located TARs successfully. Here, we present a simple adaptation of the off-the-shelf neural network RetinaNet that is designed to identify the presence of TARs at a 50-m scale. Once trained, the network was able to identify TARs with high precision (92.9%). Our model also shows promising results for applications to other surficial features like ripples and polygonal terrain. In the future, we hope to apply this model more broadly and generate a large database of TAR distributions on Mars. Full article
(This article belongs to the Special Issue Deep Learning for Remote Sensing Data)
Show Figures

Graphical abstract

Back to TopTop