Three-Order Tucker Decomposition and Reconstruction Detector for Unsupervised Hyperspectral Change Detection

Hou, Zengfu; Li, Wei; Tao, Ran; Du, Qian

doi:10.1109/jstars.2021.3088438

Cited by 52 publications

(14 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, four different state-of-theart HSI-CD methods were used to compare the effectiveness of the proposed approach. These methods are included GetNet [24], TDRD [40], PTCD [41], and ViT [42] (standard vision transformed based method) that is applied in an unsupervised manner and without any sample dataset. It is worth noting that TDRD and PTCD require threshold selection that uses the K-Means algorithm for thresholding.…”

Section: Resultsmentioning

confidence: 99%

SSViT-HCD: A Spatial–Spectral Convolutional Vision Transformer for Hyperspectral Change Detection

Shafique

Seydi

Alipour-Fard

et al. 2023

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

View full text Add to dashboard Cite

In recent decades, the wide use of deep learningbased methods has consistently improved the performance of remote sensing images and is widely used for hyperspectral change detection (HCD) tasks. However, most of the existing HCD method is based on the convolutional neural network (CNN), which shows limitations in long-range dependencies and also cannot mine sequence features well. The CD performance still has margins for improvement. In this study, inspired by the excellent performance of transformers in computer vision and which has shown a significant ability to model global dependencies to attenuate the loss of long-range information, we built a hybrid spatial-spectral convolutional vision transformer (SSViT) for HCD. Our proposed method combines the merits of CNN and transformer to fulfill effective and efficient HCD. This study focused on highly reliable pseudo-sample data generation by selection scenario. To generate a pseudo sample, we have used different methods: (1) we predict change and no-change areas by using Euclidean distance, (2) thresholding by Chan-Vese segmentation method for determining change and no-change pixels for intensity maps, (3) sorting of change and no-change pixels, and (4) selection of the minimum value of initial no-change pixels as pseudo change sample data, in addition to, choosing the maximum intensity value for change candidate pixels as change sample data. The highly reliable change pixels were selected, and then pseudo-training data was used to train the SSViT model. At last, the change map is generated by training the SSViT network based on pseudo-training data. The performance of the SSViT model is evaluated for real-world hyperspectral (HS) datasets with different change landcover types. Furthermore, a new series of HS images is introduced for CD purposes. The results of CD show that the HS images have a high potential for detecting subtle changes. The experimental results demonstrate that the proposed SSViT could outperform the advanced HCD methods.

show abstract

Section: Resultsmentioning

confidence: 99%

SSViT-HCD: A Spatial–Spectral Convolutional Vision Transformer for Hyperspectral Change Detection

Shafique

Seydi

Alipour-Fard

et al. 2023

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

View full text Add to dashboard Cite

show abstract

“…Self-supervised Tensor Network (SSTN) [3] combines self-learning and tensor factorization network to a model, which extracts low-rank features to improve the results of CD. Three-order Tucker Decomposition and Reconstruction Detector (TDRD) [4] uses tucker decomposition to remove impurity signals of HIS and it designs a singular value energy accumulation strategy to obtain principal components numbers of various factor matrixes.…”

Section: Related Workmentioning

confidence: 99%

“…We choose Change Vector Analysis (CVA) [1], Subspace-based CD method (SCD) [2] as baseline and Self-supervised Tensor Network (SSTN) [3], Three-order Tucker Decomposition and Reconstruction Detector (TDRD) [4] as SOTA to compare with our method. CVA is a classic unsupervised change detection method using polar coordinates to represent spectral change vectors and EM optimization to learn change features.…”

Section: Experiments Settingmentioning

confidence: 99%

An unsupervised symmetric tensor network for change detection in multitemporal hyperspectral images

Liang

Chen²

2023

Earth and Space: From Infrared to Terahertz (ESIT 2022)

View full text Add to dashboard Cite

Since Hyperspectral images (HSIs) contain a large amount of spectral information, they can provide detailed spectral information and enable accurate CD. However, the spectral heterogeneity of HSIs may lead to false alarms which will reduce detection accuracy. Additionally, it is difficult to collect and annotate pixel-level labels for CD in HSIs. Therefore, we propose an unsupervised symmetric tensor network (USTN) for HSIs CD. We design a novel multidimensional symmetric tensor framework to solve the problem of high-dimensional data processing. Furthermore, the framework integrates a spatial edge loss to preserve detailed spectral-spatial information. Finally, we use feature fusion to suppress the invariant components (i.e., the background) and highlight the variant components (i.e., temporal changes). Experiments on two sets of multitemporal HSIs, Hermiston and Bay Area, demonstrate the effectiveness of USTN for binary change detection.

show abstract

“…In addition, because of the 3-D characteristics of the HSIs, the three-order tensors are also utilized to reconstruct the background HSI [37]. The tensor decomposition-based detector (TenB) is applied to eliminate the background, and highlights the anomalies.…”

Section: Introductionmentioning

confidence: 99%

Dynamic Negative Sampling Autoencoder for Hyperspectral Anomaly Detection

Wang

Sun

Xia

et al. 2022

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

View full text Add to dashboard Cite

Hyperspectral anomaly detection (HAD) aims at detecting the anomalies without any prerequisite information, which gains lots of attention in recent years. Most of existing detectors locate the anomalies by eliminating the background. The background is usually reconstructed by utilizing only the global and local homogenous attribute, no matter the matrix decompositionbased or the deep learning-based methods. In this article, a dynamic negative sampling autoencoder is proposed for the hyperspectral anomaly detection (DNA-HAD). Some pixels are randomly selected and altered as negative samples. Both the rest original pixels and the altered negative samples are sent into the network. An adaptive adjusted loss function is designed to suppress the reconstruction error for the original pixels, and to enlarge the error for the negative samples. Meanwhile, skip connection is designed to ensure features at both shallow levels and deep levels being utilized for the reconstructing process. In this way, by importing some negative samples in the reconstructing process, the proposed DNA-HAD is not only robust to reconstruct the background but also sensitive to detect the anomalies. Experiments on six hyperspectral imagery which are captured by different sensors have demonstrated the effectiveness of the proposed method.

show abstract

Three-Order Tucker Decomposition and Reconstruction Detector for Unsupervised Hyperspectral Change Detection

Cited by 52 publications

References 46 publications

SSViT-HCD: A Spatial–Spectral Convolutional Vision Transformer for Hyperspectral Change Detection

SSViT-HCD: A Spatial–Spectral Convolutional Vision Transformer for Hyperspectral Change Detection

An unsupervised symmetric tensor network for change detection in multitemporal hyperspectral images

Dynamic Negative Sampling Autoencoder for Hyperspectral Anomaly Detection

Contact Info

Product

Resources

About