Local Manifold-Based Sparse Discriminant Learning for Feature Extraction of Hyperspectral Image

Duan, Yule; Hong, Huixiao; Li, Zhengying; Tang, Yuxiao

doi:10.1109/tcyb.2020.2977461

Cited by 42 publications

(12 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These HSR land-cover datasets have all promoted the development of semantic segmentation, and many variants of FCNs [19] have been evaluated [7,10,11,46]. Recently, some UDA methods have been developed from the combination of two public datasets [50].…”

Section: Land-cover Semantic Segmentation Datasetsmentioning

confidence: 99%

LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation

Wang

Zheng

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep learning approaches have shown promising results in remote sensing high spatial resolution (HSR) land-cover mapping. However, urban and rural scenes can show completely different geographical landscapes, and the inadequate generalizability of these algorithms hinders city-level or national-level mapping. Most of the existing HSR land-cover datasets mainly promote the research of learning semantic representation, thereby ignoring the model transferability. In this paper, we introduce the Land-cOVEr Domain Adaptive semantic segmentation (LoveDA) dataset to advance semantic and transferable learning. The LoveDA dataset contains 5987 HSR images with 166768 annotated objects from three different cities. Compared to the existing datasets, the LoveDA dataset encompasses two domains (urban and rural), which brings considerable challenges due to the: 1) multi-scale objects; 2) complex background samples; and 3) inconsistent class distributions. The LoveDA dataset is suitable for both land-cover semantic segmentation and unsupervised domain adaptation (UDA) tasks. Accordingly, we benchmarked the LoveDA dataset on eleven semantic segmentation methods and eight UDA methods. Some exploratory studies including multi-scale architectures and strategies, additional background supervision, and pseudo-label analysis were also carried out to address these challenges. The code and data are available at https://github.com/Junjue-Wang/LoveDA.

show abstract

Section: Land-cover Semantic Segmentation Datasetsmentioning

confidence: 99%

LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation

Wang

Zheng

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…where D wh (x 1 , x 2 ) is the weighted Hamming distance described in Equation (5). In the revised Hamming distance if there are two absorption occurs within the window, they are considered as the same absorption and the Hamming distance of the windowed segmentation is calculated as 1 rather than 2.…”

Section: Improved Absorption Detectionmentioning

confidence: 99%

“…Apparently, these features should be as effective as possible to include the fundamental characteristics in line with each type of the materials. In the past two decades, many feature representations have been investigated, such as the spectra curves [3], the subsets of spectral bands [4], the features based on manifold learning [5], the spatial-spectral features [6] etc.…”

Section: Introductionmentioning

confidence: 99%

Matching hyperspectral absorptions by weighted hamming distance

Guo

2021

Electronics Letters

View full text Add to dashboard Cite

To analyse and compare hyperspectral signatures, features extraction and matching are two key issues. In this letter, hyperspectral absorption features and the corresponding matching algorithm are discussed. First, an absorption detection method is applied to catch all necessary spectral absorptions with improved reliability. Then, a weighted Hamming distance is proposed to match the binary absorption-features. Next, an elastic matching scheme is designed to classify the hyperspectral data. Experiments of classification are carried out on six classes of vegetation from the Salinas data-set. Results show that the proposed method not only increased the overall classification accuracy to 73.13% from back propagation neural network's 71.86% and support vector machine's 73.06%, but also improved the error distributions among different classes.

show abstract

“…In the early stage, various kinds of feature extraction (FE) methods have been developed for processing high-dimensional data [9], [10]. As a common finite element method, graph learning method obtains embedded features by projecting high-dimensional data into a potential low-dimensional space [11], [12]. For example, a local neighborhood graph is constructed by local constrained manifold structure collaborative preserving embedding (LMSCPE) [13] to improve the aggregation of HSI data, and collaborative representation theory is adopted to effectively characterize the correlation between data pairs.…”

Section: Introductionmentioning

confidence: 99%

Aggregated-Attention Transformation Network for Hyperspectral Image Classification

Hong

2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

Convolutional neural networks (CNNs) have achieved great success in hyperspectral image (HSI) classification. However, most CNNs require a substantial amount of manually labeled training data to yield good performances, and the expert annotation of HSI data is labor-intensive, but existing CNNs fail to effectively extract the joint spatial-spectral features from a small training set. To tackle this issue, inspired by the attention mechanism, an aggregated-attention transformation network (AATN) was proposed for HSI classification. This network takes advantage of the split-transform-merge strategy to change the number of the cardinality (the size of transformation set) for improving generalization ability with a small number of training samples while alleviating the overfitting phenomenon. Moreover, each aggregation attention module is designed to explore the attention across spatial-spectral features of different receptive fields, and interactively guide the recalibration of discriminative features. By retaining the uniform architecture, AATN can be used for HSI classification tasks without introducing excessive computational costs, and the proposed framework can learn representations with limited training samples due to the effectiveness and practicality of the attention-driven features. The proposed AATN obtains competitive performance compared with some state-of-the-art methods in qualitative and quantitative experiments on four benchmark HSI datasets.

show abstract

Local Manifold-Based Sparse Discriminant Learning for Feature Extraction of Hyperspectral Image

Cited by 42 publications

References 35 publications

LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation

LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation

Matching hyperspectral absorptions by weighted hamming distance

Aggregated-Attention Transformation Network for Hyperspectral Image Classification

Contact Info

Product

Resources

About