Cargando…

RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification

While growing instruments generate more and more airborne or satellite images, the bottleneck in remote sensing (RS) scene classification has shifted from data limits toward a lack of ground truth samples. There are still many challenges when we are facing unknown environments, especially those with...

Descripción completa

Detalles Bibliográficos
Autores principales: Zhang, Pei, Li, Ying, Wang, Dong, Wang, Jiyue
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7956409/
https://www.ncbi.nlm.nih.gov/pubmed/33668138
http://dx.doi.org/10.3390/s21051566
_version_ 1783664428686770176
author Zhang, Pei
Li, Ying
Wang, Dong
Wang, Jiyue
author_facet Zhang, Pei
Li, Ying
Wang, Dong
Wang, Jiyue
author_sort Zhang, Pei
collection PubMed
description While growing instruments generate more and more airborne or satellite images, the bottleneck in remote sensing (RS) scene classification has shifted from data limits toward a lack of ground truth samples. There are still many challenges when we are facing unknown environments, especially those with insufficient training data. Few-shot classification offers a different picture under the umbrella of meta-learning: digging rich knowledge from a few data are possible. In this work, we propose a method named RS-SSKD for few-shot RS scene classification from a perspective of generating powerful representation for the downstream meta-learner. Firstly, we propose a novel two-branch network that takes three pairs of original-transformed images as inputs and incorporates Class Activation Maps (CAMs) to drive the network mining, the most relevant category-specific region. This strategy ensures that the network generates discriminative embeddings. Secondly, we set a round of self-knowledge distillation to prevent overfitting and boost the performance. Our experiments show that the proposed method surpasses current state-of-the-art approaches on two challenging RS scene datasets: NWPU-RESISC45 and RSD46-WHU. Finally, we conduct various ablation experiments to investigate the effect of each component of the proposed method and analyze the training time of state-of-the-art methods and ours.
format Online
Article
Text
id pubmed-7956409
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-79564092021-03-16 RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification Zhang, Pei Li, Ying Wang, Dong Wang, Jiyue Sensors (Basel) Article While growing instruments generate more and more airborne or satellite images, the bottleneck in remote sensing (RS) scene classification has shifted from data limits toward a lack of ground truth samples. There are still many challenges when we are facing unknown environments, especially those with insufficient training data. Few-shot classification offers a different picture under the umbrella of meta-learning: digging rich knowledge from a few data are possible. In this work, we propose a method named RS-SSKD for few-shot RS scene classification from a perspective of generating powerful representation for the downstream meta-learner. Firstly, we propose a novel two-branch network that takes three pairs of original-transformed images as inputs and incorporates Class Activation Maps (CAMs) to drive the network mining, the most relevant category-specific region. This strategy ensures that the network generates discriminative embeddings. Secondly, we set a round of self-knowledge distillation to prevent overfitting and boost the performance. Our experiments show that the proposed method surpasses current state-of-the-art approaches on two challenging RS scene datasets: NWPU-RESISC45 and RSD46-WHU. Finally, we conduct various ablation experiments to investigate the effect of each component of the proposed method and analyze the training time of state-of-the-art methods and ours. MDPI 2021-02-24 /pmc/articles/PMC7956409/ /pubmed/33668138 http://dx.doi.org/10.3390/s21051566 Text en © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Zhang, Pei
Li, Ying
Wang, Dong
Wang, Jiyue
RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification
title RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification
title_full RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification
title_fullStr RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification
title_full_unstemmed RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification
title_short RS-SSKD: Self-Supervision Equipped with Knowledge Distillation for Few-Shot Remote Sensing Scene Classification
title_sort rs-sskd: self-supervision equipped with knowledge distillation for few-shot remote sensing scene classification
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7956409/
https://www.ncbi.nlm.nih.gov/pubmed/33668138
http://dx.doi.org/10.3390/s21051566
work_keys_str_mv AT zhangpei rssskdselfsupervisionequippedwithknowledgedistillationforfewshotremotesensingsceneclassification
AT liying rssskdselfsupervisionequippedwithknowledgedistillationforfewshotremotesensingsceneclassification
AT wangdong rssskdselfsupervisionequippedwithknowledgedistillationforfewshotremotesensingsceneclassification
AT wangjiyue rssskdselfsupervisionequippedwithknowledgedistillationforfewshotremotesensingsceneclassification