Abstract:Hyperspectral unmixing (HU) aims to decompose mixed pixels into a set of endmembers and corresponding abundances. Deep learning-based HU methods are currently a hot research topic, but most existing unmixing methods still rely on per-pixel training or employ convolutional neural networks (CNNs), which overlook the non-local correlations of materials and spectral characteristics. Furthermore, current research mainly focuses on linear mixing models, which limits the feature extraction capability of deep encoders and further improvement in unmixing accuracy. In this paper, we propose a nonlinear unmixing network capable of extracting global spatial-spectral features. The network is designed based on an autoencoder architecture, where a dual-stream CNNs is employed in the encoder to separately extract spectral and local spatial information. The extracted features are then fused together to form a more complete representation of the input data. Subsequently, a linear projection-based multi-head self-attention mechanism is applied to capture global contextual information, allowing for comprehensive spatial information extraction while maintaining lightweight computation. To achieve better reconstruction performance, a model-free nonlinear mixing approach is adopted to enhance the model's universality, with the mixing model learned entirely from the data. Additionally, an initialization method based on endmember bundles is utilized to reduce interference from outliers and noise. Comparative results on real datasets against several state-of-the-art unmixing methods demonstrate the superior of the proposed approach.

Nonlinear Unmixing of Hyperspectral Data via Deep Autoencoder Networks

Hyperspectral Unmixing via Deep Autoencoder Networks for a Generalized Linear-Mixture/Nonlinear-Fluctuation Model

Hyperspectral Unmixing for Additive Nonlinear Models With a 3-D-CNN Autoencoder Network

Nonlinear Hyperspectral Unmixing based on Multilinear Mixing Model using Convolutional Autoencoders

LSTM-DNN Based Autoencoder Network for Nonlinear Hyperspectral Image Unmixing

Nonlinear Unmixing of Hyperspectral Data with Partially Linear Least-Squares Support Vector Regression

Hyperspectral Unmixing Network Accounting for Spectral Variability Based on a Modified Scaled and a Perturbed Linear Mixing Model

Autoencoder Network for Hyperspectral Unmixing with Adaptive Abundance Smoothing

Unrolling Plug-and-Play Network for Hyperspectral Unmixing

Unsupervised Nonlinear Hyperspectral Unmixing Based on an Extended Multilinear Mixing Model-Inspired Dual-Stream Network

Deep Auto-Encoder Network for Hyperspectral Image Unmixing.

Nonlinear Unmixing Of Hyperspectral Data Based On A Linear-Mixture/Nonlinear-Fluctuation Model

Hopfield Neural Network Approach for Supervised Nonlinear Spectral Unmixing

NMF-SAE: an Interpretable Sparse Autoencoder for Hyperspectral Unmixing

Nonlinear unmixing of hyperspectral images based on multi-kernel learning

Hyperspectral Unmixing Via Wavelet Based Autoencoder Network

Integration of Physics-Based and Data-Driven Models for Hyperspectral Image Unmixing

Deep Nonlinear Hyperspectral Unmixing Using Multi-task Learning

DAEN: Deep Autoencoder Networks for Hyperspectral Unmixing

A Global Spatial-Spectral Feature Fused Autoencoder for Nonlinear Hyperspectral Unmixing

Deep Spectral Unmixing Framework Via 3D Denoising Convolutional Autoencoder