Please use this identifier to cite or link to this item: http://dx.doi.org/10.14279/depositonce-11036
For citation please use:
Main Title: High-Rankness Regularized Semi-Supervised Deep Metric Learning for Remote Sensing Imagery
Author(s): Kang, Jian
Fernández-Beltrán, Rubén
Ye, Zhen
Tong, Xiaohua
Ghamisi, Pedram
Plaza, Antonio
Type: Article
Language Code: en
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 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.
URI: https://depositonce.tu-berlin.de/handle/11303/12162
http://dx.doi.org/10.14279/depositonce-11036
Issue Date: 12-Aug-2020
Date Available: 9-Dec-2020
DDC Class: 620 Ingenieurwissenschaften und zugeordnete Tätigkeiten
Subject(s): deep metric learning
remote sensing
image characterization
semi-supervised learning
Sponsor/Funder: EC/H2020/734541/EU/TOOLS FOR MAPPING HUMAN EXPOSURE TO RISKY ENVIRONMENTAL CONDITIONS BY MEANS OF GROUND AND EARTH OBSERVATION DATA/EOXPOSURE
License: https://creativecommons.org/licenses/by/4.0/
Journal Title: Remote Sensing
Publisher: MDPI
Publisher Place: Basel
Volume: 12
Issue: 16
Article Number: 2603
Publisher DOI: 10.3390/rs12162603
EISSN: 2072-4292
Appears in Collections:FG Remote Sensing Image Analysis Group » Publications

Files in This Item:
remotesensing-12-02603-v2.pdf
Format: Adobe PDF | Size: 4.58 MB
DownloadShow Preview
Thumbnail

Item Export Bar

This item is licensed under a Creative Commons License Creative Commons