Skip to main content

A Self-Training Hierarchical Prototype-based Ensemble Framework for Remote Sensing Scene Classification

Gu, Xiaowei, Zhang, Ce, Shen, Qiang, Han, Jungong, Angelov, Plamen, Atkinson, Peter (2022) A Self-Training Hierarchical Prototype-based Ensemble Framework for Remote Sensing Scene Classification. Information Fusion, 80 . pp. 179-204. ISSN 1566-2535. (doi:10.1016/j.inffus.2021.11.014) (Access to this publication is currently restricted. You may be able to access a copy if URLs are provided) (KAR id:91586)

PDF Author's Accepted Manuscript
Language: English

Restricted to Repository staff only until 17 November 2022.

Contact us about this Publication
[thumbnail of STHPensemble_final.pdf]
Official URL
https://doi.org/10.1016/j.inffus.2021.11.014

Abstract

Remote sensing scene classification plays a critical role in a wide range of real-world applications. Technically, however, scene classification is an extremely challenging task due to the huge complexity in remotely sensed scenes, and the difficulty in acquiring labelled data for model training such as supervised deep learning. To tackle these issues, a novel semi-supervised ensemble framework is proposed here using the self-training hierarchical prototype-based classifier as the base learner for chunk-by-chunk prediction. The framework has the ability to build a powerful ensemble model from both labelled and unlabelled images with minimum supervision. Different feature descriptors are employed in the proposed ensemble framework to offer multiple independent views of images. Thus, the diversity of base learners is guaranteed for ensemble classification. To further increase the overall accuracy, a novel cross-checking strategy was introduced to enable the base learners to exchange pseudo-labelling information during the self-training process, and maximize the correctness of pseudo-labels assigned to unlabelled images. Extensive numerical experiments on popular benchmark remote sensing scenes demonstrated the effectiveness of the proposed ensemble framework, especially where the number of labelled images available is limited. For example, the classification accuracy achieved on the OPTIMAL-31, PatternNet and RSI-CB256 datasets was up to 99.91%, 98. 67% and 99.07% with only 40% of the image sets used as labelled training images, surpassing or at least on par with mainstream benchmark approaches trained with double the number of labelled images.

Item Type: Article
DOI/Identification number: 10.1016/j.inffus.2021.11.014
Uncontrolled keywords: self-training; pseudo-labelling; prototypes; remote sensing; scene classification
Subjects: Q Science > QA Mathematics (inc Computing science)
Divisions: Divisions > Division of Computing, Engineering and Mathematical Sciences > School of Computing
Depositing User: Xiaowei Gu
Date Deposited: 15 Nov 2021 14:31 UTC
Last Modified: 13 Jan 2022 16:32 UTC
Resource URI: https://kar.kent.ac.uk/id/eprint/91586 (The current URI for this page, for reference purposes)
  • Depositors only (login required):