Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification

Haikel Alhichri; Naif Alajlan; Yakoub Bazi; Timon Rabczuk

doi:10.1109/EIT.2018.8500107

Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification

Haikel Alhichri, Naif Alajlan, Yakoub Bazi, Timon Rabczuk

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

8 Citations (Scopus)

Abstract

In recent years the problem of scene classification in remote sensing has attracted a considerable amount of attention. Solution for this important problem based on deep convolutional neural networks (CNN) are currently state-of-the-art. So far all CNNs used images of fixed size (typically 224× 224 which commonly used in other fields of computer vision). In this paper, we propose a multi-scale deep CNN architecture that can accept variable image sizes. We achieve this by using multiple CNN, that share some or all parameters, followed by a merge layer, fully connected layers, and finally a softmax layer for classification. In each epoch we train the network with a batch of images of all scales. We have implemented this architecture using three SqueezeNet CNNs trained on three different scales of scene images. Then we carried out experiments on three well know datasets, namely UC Merced, KSA, and AID datasets. Preliminary results show that this multi-scale CNN do converge just as the traditional single-scale training, and leads to better testing accuracy.

Original language	English
Title of host publication	2018 IEEE International Conference on Electro/Information Technology, EIT 2018
Publisher	IEEE Computer Society
Pages	113-117
Number of pages	5
Volume	2018-May
ISBN (Electronic)	9781538653982
DOIs	https://doi.org/10.1109/EIT.2018.8500107
Publication status	Published - 2018 Oct 18
Externally published	Yes
Event	2018 IEEE International Conference on Electro/Information Technology, EIT 2018 - Rochester, United States Duration: 2018 May 3 → 2018 May 5

Other

Other	2018 IEEE International Conference on Electro/Information Technology, EIT 2018
Country	United States
City	Rochester
Period	18/5/3 → 18/5/5

Keywords

Convolutional neural networks (CNN)
Deep learning
Deep neural networks
Remote sensing
Scene classification

ASJC Scopus subject areas

Computer Science Applications
Information Systems
Control and Systems Engineering
Electrical and Electronic Engineering

Access to Document

10.1109/EIT.2018.8500107

Fingerprint Dive into the research topics of 'Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification'. Together they form a unique fingerprint.

Cite this

Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification. / Alhichri, Haikel; Alajlan, Naif; Bazi, Yakoub; Rabczuk, Timon.

2018 IEEE International Conference on Electro/Information Technology, EIT 2018. Vol. 2018-May IEEE Computer Society, 2018. p. 113-117 8500107.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Alhichri, H, Alajlan, N, Bazi, Y & Rabczuk, T 2018, Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification. in 2018 IEEE International Conference on Electro/Information Technology, EIT 2018. vol. 2018-May, 8500107, IEEE Computer Society, pp. 113-117, 2018 IEEE International Conference on Electro/Information Technology, EIT 2018, Rochester, United States, 18/5/3. https://doi.org/10.1109/EIT.2018.8500107

@inproceedings{c7bc9f273fdc4056922a93203dde3941,

title = "Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification",

abstract = "In recent years the problem of scene classification in remote sensing has attracted a considerable amount of attention. Solution for this important problem based on deep convolutional neural networks (CNN) are currently state-of-the-art. So far all CNNs used images of fixed size (typically 224× 224 which commonly used in other fields of computer vision). In this paper, we propose a multi-scale deep CNN architecture that can accept variable image sizes. We achieve this by using multiple CNN, that share some or all parameters, followed by a merge layer, fully connected layers, and finally a softmax layer for classification. In each epoch we train the network with a batch of images of all scales. We have implemented this architecture using three SqueezeNet CNNs trained on three different scales of scene images. Then we carried out experiments on three well know datasets, namely UC Merced, KSA, and AID datasets. Preliminary results show that this multi-scale CNN do converge just as the traditional single-scale training, and leads to better testing accuracy.",

keywords = "Convolutional neural networks (CNN), Deep learning, Deep neural networks, Remote sensing, Scene classification",

author = "Haikel Alhichri and Naif Alajlan and Yakoub Bazi and Timon Rabczuk",

year = "2018",

month = oct,

day = "18",

doi = "10.1109/EIT.2018.8500107",

language = "English",

volume = "2018-May",

pages = "113--117",

booktitle = "2018 IEEE International Conference on Electro/Information Technology, EIT 2018",

publisher = "IEEE Computer Society",

note = "2018 IEEE International Conference on Electro/Information Technology, EIT 2018 ; Conference date: 03-05-2018 Through 05-05-2018",

}

TY - GEN

T1 - Multi-Scale Convolutional Neural Network for Remote Sensing Scene Classification

AU - Alhichri, Haikel

AU - Alajlan, Naif

AU - Bazi, Yakoub

AU - Rabczuk, Timon

PY - 2018/10/18

Y1 - 2018/10/18

N2 - In recent years the problem of scene classification in remote sensing has attracted a considerable amount of attention. Solution for this important problem based on deep convolutional neural networks (CNN) are currently state-of-the-art. So far all CNNs used images of fixed size (typically 224× 224 which commonly used in other fields of computer vision). In this paper, we propose a multi-scale deep CNN architecture that can accept variable image sizes. We achieve this by using multiple CNN, that share some or all parameters, followed by a merge layer, fully connected layers, and finally a softmax layer for classification. In each epoch we train the network with a batch of images of all scales. We have implemented this architecture using three SqueezeNet CNNs trained on three different scales of scene images. Then we carried out experiments on three well know datasets, namely UC Merced, KSA, and AID datasets. Preliminary results show that this multi-scale CNN do converge just as the traditional single-scale training, and leads to better testing accuracy.

AB - In recent years the problem of scene classification in remote sensing has attracted a considerable amount of attention. Solution for this important problem based on deep convolutional neural networks (CNN) are currently state-of-the-art. So far all CNNs used images of fixed size (typically 224× 224 which commonly used in other fields of computer vision). In this paper, we propose a multi-scale deep CNN architecture that can accept variable image sizes. We achieve this by using multiple CNN, that share some or all parameters, followed by a merge layer, fully connected layers, and finally a softmax layer for classification. In each epoch we train the network with a batch of images of all scales. We have implemented this architecture using three SqueezeNet CNNs trained on three different scales of scene images. Then we carried out experiments on three well know datasets, namely UC Merced, KSA, and AID datasets. Preliminary results show that this multi-scale CNN do converge just as the traditional single-scale training, and leads to better testing accuracy.

KW - Convolutional neural networks (CNN)

KW - Deep learning

KW - Deep neural networks

KW - Remote sensing

KW - Scene classification

UR - http://www.scopus.com/inward/record.url?scp=85057107188&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057107188&partnerID=8YFLogxK

U2 - 10.1109/EIT.2018.8500107

DO - 10.1109/EIT.2018.8500107

M3 - Conference contribution

AN - SCOPUS:85057107188

VL - 2018-May

SP - 113

EP - 117

BT - 2018 IEEE International Conference on Electro/Information Technology, EIT 2018

PB - IEEE Computer Society

T2 - 2018 IEEE International Conference on Electro/Information Technology, EIT 2018

Y2 - 3 May 2018 through 5 May 2018

ER -