2021
DOI: 10.3390/rs13040729
|View full text |Cite
|
Sign up to set email alerts
|

3DeepM: An Ad Hoc Architecture Based on Deep Learning Methods for Multispectral Image Classification

Abstract: Current predefined architectures for deep learning are computationally very heavy and use tens of millions of parameters. Thus, computational costs may be prohibitive for many experimental or technological setups. We developed an ad hoc architecture for the classification of multispectral images using deep learning techniques. The architecture, called 3DeepM, is composed of 3D filter banks especially designed for the extraction of spatial-spectral features in multichannel images. The new architecture has been … Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
4
0
1

Year Published

2022
2022
2024
2024

Publication Types

Select...
8
1

Relationship

1
8

Authors

Journals

citations
Cited by 12 publications
(5 citation statements)
references
References 47 publications
0
4
0
1
Order By: Relevance
“…The contrast and brightness do change the pixel values, but its range was limited to minimize any possible hindrance in the learning process of the classifiers. The precise values of the transformation ranges are identical to those described in the literature [ 33 ].…”
Section: Potential Usage Of Datasetmentioning
confidence: 81%
See 1 more Smart Citation
“…The contrast and brightness do change the pixel values, but its range was limited to minimize any possible hindrance in the learning process of the classifiers. The precise values of the transformation ranges are identical to those described in the literature [ 33 ].…”
Section: Potential Usage Of Datasetmentioning
confidence: 81%
“…The 3D-CNN architecture employs two 3-dimensional convolutional blocks with convolutional layers that use dilated 3-dimensional kernels [ 33 ]. It also uses mean and average 3-dimensional pooling layers between these blocks and a total of 2 dense layers.…”
Section: Potential Usage Of Datasetmentioning
confidence: 99%
“…One of the common approaches for multispectral image classification is to use 3D convolutional neural networks (CNNs) that can capture both spatial and spectral features from the input images. [30] proposed an ad hoc architecture called 3DeepM that consists of 3D filter banks specially designed to extract spatial spectral features in multichannel images. The authors applied their method to several aerial and satellite image datasets and achieved competitive results.…”
Section: Introductionmentioning
confidence: 99%
“…MSI super-resolution through deep learning has become a promising method to recover additional spectral information from RGB images without using more expensive MSI or HSI hardware [6]. MSIs have been shown to be superior in classification tasks compared to corresponding RGB images [7,8]. Optimal waveband selection has been shown to improve the performances of model predictions [9,10].…”
Section: Introductionmentioning
confidence: 99%