Semisupervised Hypergraph Discriminant Learning for Dimensionality Reduction of Hyperspectral Image

Luo, Fulin; Guo, Tan; Lin, Zhiping; Ren, Jinchang; Zhou, Xiaocheng

doi:10.1109/jstars.2020.3011431

Cited by 14 publications

(4 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further to validate the performance of the proposed model, different dimensionality reduction techniques are compared with the existing methods like Enhanced Hybrid-Graph Discriminant Learning (EHGDL), local geometric structure Fisher analysis (LGSFA), Discriminant hyper-Laplacian projection (DHLP), Group-based tensor model (GBTM), and Lower rank tensor approximation (LRTA) methods. The results are obtained from Luo et al [40] and An et al [41] research works that perform dimensionality reduction in Hyperspectral images. The dataset which was used in the proposed work was used in the existing methods so that the results are directly compared with the proposed model results given in Table 3.…”

Section: Resultsmentioning

confidence: 99%

Dimensionality Reduction Using Optimized Self-Organized Map Technique for Hyperspectral Image Classification

Srinivasan¹,

Rajakumar²

2023

Computer Systems Science and Engineering

View full text Add to dashboard Cite

The high dimensionalhyperspectral image classification is a challenging task due to the spectral feature vectors. The high correlation between these features and the noises greatly affects the classification performances. To overcome this, dimensionality reduction techniques are widely used. Traditional image processing applications recently propose numerous deep learning models. However, in hyperspectral image classification, the features of deep learning models are less explored. Thus, for efficient hyperspectral image classification, a depth-wise convolutional neural network is presented in this research work. To handle the dimensionality issue in the classification process, an optimized self-organized map model is employed using a water strider optimization algorithm. The network parameters of the self-organized map are optimized by the water strider optimization which reduces the dimensionality issues and enhances the classification performances. Standard datasets such as Indian Pines and the University of Pavia (UP) are considered for experimental analysis. Existing dimensionality reduction methods like Enhanced Hybrid-Graph Discriminant Learning (EHGDL), local geometric structure Fisher analysis (LGSFA), Discriminant Hyper-Laplacian projection (DHLP), Group-based tensor model (GBTM), and Lower rank tensor approximation (LRTA) methods are compared with proposed optimized SOM model. Results confirm the superior performance of the proposed model of 98.22% accuracy for the Indian pines dataset and 98.21% accuracy for the University of Pavia dataset over the existing maximum likelihood classifier, and Support vector machine (SVM).

show abstract

Section: Resultsmentioning

confidence: 99%

Dimensionality Reduction Using Optimized Self-Organized Map Technique for Hyperspectral Image Classification

Srinivasan¹,

Rajakumar²

2023

Computer Systems Science and Engineering

View full text Add to dashboard Cite

show abstract

“…The use of HSI analysis is widespread in a variety of industries, including materials analysis, precision agriculture, environmental monitoring, and surveillance [31][32][33]. The hyperspectral community's most active area of study is HSIs classification, which aims to categories every pixel in an image [34]. In figure 4 show the concept of hyperspectral imaging.…”

Section: Hyperspectral Imagingmentioning

confidence: 99%

Classification of Rethinking Hyperspectral Images using 2D and 3D CNN with Channel and Spatial Attention: A Review

Aslam¹,

Ali²,

Nawaz³

et al. 2023

J. Engg. Res. & Sci.

View full text Add to dashboard Cite

It has been demonstrated that 3D Convolutional Neural Networks (CNN) are an effective technique for classifying hyperspectral images (HSI). Conventional 3D CNNs produce too many parameters to extract the spectral-spatial properties of HSIs. A channel service module and a spatial service module are utilized to optimize characteristic maps and enhance sorting performance in order to further study discriminating characteristics. In this article, evaluate CNN's methods for hyperspectral image categorization (HSI). Examined the replacement of traditional 3D CNN with mixed feature maps by frequency to lessen spatial redundancy and expand the receptive field. Evaluates several CNN stories that use image classification algorithms, elaborating on the efficacy of these approaches or any remaining holes in methods. How do improve those gaps for better image classification?

show abstract

“…A general sequential process in a HS image classification system consists of image restoration (e.g., denoising, missing data recovery) [2]- [4], dimensionality reduction [5]- [7], spectral unmixing [8]- [10], and feature extraction [11]- [13]. Among them, feature extraction is a crucial step in HS image classification, which has received increasing attention by researchers.…”

mentioning

confidence: 99%

“…7,8, and 10 provide the obtained results for the Indian Pines, Pavia University, and Houston2013 datasets, respectively. Roughly, conventional classification models (e.g., KNN, RF, SVM) tend…”

mentioning

confidence: 99%

SpectralFormer: Rethinking Hyperspectral Image Classification with Transformers

Hong,

Han,

Yao

et al. 2021

Preprint

View full text Add to dashboard Cite

Hyperspectral (HS) images are characterized by approximately contiguous spectral information, enabling the fine identification of materials by capturing subtle spectral discrepancies. Owing to their excellent locally contextual modeling ability, convolutional neural networks (CNNs) have been proven to be a powerful feature extractor in HS image classification. However, CNNs fail to mine and represent the sequence attributes of spectral signatures well due to the limitations of their inherent network backbone. To solve this issue, we rethink HS image classification from a sequential perspective with transformers, and propose a novel backbone network called SpectralFormer. Beyond band-wise representations in classic transformers, Spec-tralFormer is capable of learning spectrally local sequence information from neighboring bands of HS images, yielding group-wise spectral embeddings. More significantly, to reduce the possibility of losing valuable information in the layer-wise propagation process, we devise a cross-layer skip connection to convey memory-like components from shallow to deep layers by adaptively learning to fuse "soft" residuals across layers. It is worth noting that the proposed SpectralFormer is a highly flexible backbone network, which can be applicable to both pixel-and patch-wise inputs. We evaluate the classification performance of the proposed SpectralFormer on three HS datasets by conducting extensive experiments, showing the superiority over classic transformers and achieving a significant improvement in comparison with state-of-the-art backbone networks. The codes of this work will be available at https://sites.google.com/view/danfeng-hong for the sake of reproducibility.

show abstract

Semisupervised Hypergraph Discriminant Learning for Dimensionality Reduction of Hyperspectral Image

Cited by 14 publications

References 43 publications

Dimensionality Reduction Using Optimized Self-Organized Map Technique for Hyperspectral Image Classification

Dimensionality Reduction Using Optimized Self-Organized Map Technique for Hyperspectral Image Classification

Classification of Rethinking Hyperspectral Images using 2D and 3D CNN with Channel and Spatial Attention: A Review

SpectralFormer: Rethinking Hyperspectral Image Classification with Transformers

Contact Info

Product

Resources

About