Residual Dense Autoencoder Network for Nonlinear Hyperspectral Unmixing

Yang, Xu; Chen, Jianguo; Wang, Chengbin; Chen, Zihao

doi:10.1109/jstars.2022.3188565

Cited by 14 publications

(2 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It was trained by a loss function including a quadratic term based on the Hapke model, the reconstruction error of the reflectances, and a minimum volume total variation (TV) term. A residual dense autoencoder network is constructed in [38] for nonlinear decomposition in multiple scattering scenes. And a new type of deep automatic encoder network is designed based on the generalized bilinear model in [39].…”

Section: Model-guided DL Methodsmentioning

confidence: 99%

IVIU-Net: Implicit Variable Iterative Unrolling Network for Hyperspectral Sparse Unmixing

Shao

Liu

Xiao

2023

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

View full text Add to dashboard Cite

At present, an emerging technique called the algorithm unrolling approach has attracted wide attention, because it is capable of developing efficient and interpretable layers to eliminate the black-box nature of deep learning (DL). In this paper, inspired by the sparse unmixing model, we propose a model-driven DL approach, namely, an implicit variable iterative unrolling network (IVIU-Net). First of all, the unmixing performance and adaptive ability of the model are enhanced by introducing learnable parameters into the sparse unmixing algorithm. Then, a specific spatial convolution module is integrated into the network to promote the smoothness of the latent abundance map. Finally, a comprehensive loss function with three terms such as average spectral angle distance, HSIs reconstruction error, and spectral information divergence, is presented to train the IVIU-Net in an unsupervised way. Compared to the unmixing results of most existing data-driven DL algorithms, our network has significant advantages in two folds: 1) it is able to achieve better stability instead of relying heavily on the endmember initialization results; 2) it has better interpretability and robustness in the unmixing procedure. Experimental results on synthetic and real data show that the proposed network outperforms the state-of-the-art in terms of better convergence, faster unmixing speed as well as better accuracy.

show abstract

Section: Model-guided DL Methodsmentioning

confidence: 99%

IVIU-Net: Implicit Variable Iterative Unrolling Network for Hyperspectral Sparse Unmixing

Shao

Liu

Xiao

2023

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

View full text Add to dashboard Cite

show abstract

“…Two main parts make up the DualMAF module: channel attention methods and atrous convolution with various expansion rates. Rich feature information can be captured by the atrous convolution method, which can increase the convolution kernel's receptive field without changing the computational parameters [28]. Moreover, channel attention is added to reinforce the interaction between several sequences in order to enhance the feature recognition ability.…”

Section: Dual-path Multi-scale Attention Fusion Modulementioning

confidence: 99%

Enhanced Lung Nodule Segmentation using Dung Beetle Optimization based LNS-DualMAGNet Model

2024

Int. Res. J. multidiscip. Technovation

View full text Add to dashboard Cite

The study's focus is on lung nodules, which are frequently connected to lung cancer, the world's most common cause of cancer-related deaths. In clinical practice, a timely and precise diagnosis of these nodules is essential, albeit difficult. For diagnosis, the study used CT scans from the Lung Image Database Consortium and the LIDC-IDRI dataset. Noise reduction with a Gaussian Smoothing (GS) Filter and contrast enhancement were part of the preprocessing. With a Dual-path Multi-scale Attention Fusion Module (DualMAF) and a Multi-scale Normalized Channel Attention Module (MNCA), the study presented the LNS-DualMAGNet model for lung nodule segmentation. These modules improve interdependence across channels and semantic understanding by utilizing novel approaches such as Depthwise Separable Convolutions and attention mechanisms. For increased performance, the model also incorporates DSConv and a Resnet34 block. The Dung Beetle Optimization Algorithm (DBOA) was used for tuning the hyperparameter of the proposed classifier. Findings indicated that the proposed model performed better than the existing approaches, attaining a 0.99 accuracy and DSC, indicating its potential to enhance lung nodule segmentation for clinical diagnosis.

show abstract

Coupled Dense Convolutional Neural Networks with Autoencoder for Unsupervised Hyperspectral Super-Resolution

Lin,

Su,

et al. 2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Residual Dense Autoencoder Network for Nonlinear Hyperspectral Unmixing

Cited by 14 publications

References 69 publications

IVIU-Net: Implicit Variable Iterative Unrolling Network for Hyperspectral Sparse Unmixing

IVIU-Net: Implicit Variable Iterative Unrolling Network for Hyperspectral Sparse Unmixing

Enhanced Lung Nodule Segmentation using Dung Beetle Optimization based LNS-DualMAGNet Model

Coupled Dense Convolutional Neural Networks with Autoencoder for Unsupervised Hyperspectral Super-Resolution

Contact Info

Product

Resources

About