Abstract:Due to the superior spatial–spectral extraction capability of the convolutional neural network (CNN), CNN shows great potential in dimensionality reduction (DR) of hyperspectral images (HSIs). However, most CNN-based methods are supervised while the class labels of HSIs are limited and difficult to obtain. While a few unsupervised CNN-based methods have been proposed recently, they always focus on data reconstruction and are lacking in the exploration of discriminability which is usually the primary goal of DR. To address these issues, we propose a deep fully convolutional embedding network (DFCEN), which not only considers data reconstruction but also introduces the specific learning task of enhancing feature discriminability. DFCEN has an end-to-end symmetric network structure that is the key for unsupervised learning. Moreover, a novel objective function containing two terms—the reconstruction term and the embedding term of a specific task—is established to supervise the learning of DFCEN towards improving the completeness and discriminability of low-dimensional data. In particular, the specific task is designed to explore and preserve relationships among samples in HSIs. Besides, due to the limited training samples, inherent complexity and the presence of noise in HSIs, a preprocessing where a few noise spectral bands are removed is adopted to improve the effectiveness of unsupervised DFCEN. Experimental results on three well-known hyperspectral datasets and two classifiers illustrate that the low dimensional features of DFCEN are highly separable and DFCEN has promising classification performance compared with other DR methods.

Discriminant Collaborative Neighborhood Preserving Embedding for Hyperspectral Imagery

Dimensionality Reduction of Hyperspectral Image Based on Local Constrained Manifold Structure Collaborative Preserving Embedding

Multi-layer Cross-domain Non-negative Matrix Factorization for Cross-scene Dimension Reduction on Hyperspectral Images.

Laplacian Regularized Spatial-Aware Collaborative Graph for Discriminant Analysis of Hyperspectral Imagery

Dimension Reduction of Hyperspectral Images with Sparse Linear Discriminant Analysis

Joint Sparse Local Linear Discriminant Analysis for Feature Dimensionality Reduction of Hyperspectral Images

Spectral-Spatial Discriminant Feature Learning For Hyperspectral Image Classification

Deep Fully Convolutional Embedding Networks for Hyperspectral Images Dimensionality Reduction

Semi-supervised Enhanced Discriminative Local Constraint Preserving Projection for Dimensionality Reduction of Medical Hyperspectral Images

Collaborative Label Propagation-based Semisupervised Linear Discriminant Analysis for Hyperspectral Images

Residual Deep PCA-based Feature Extraction for Hyperspectral Image Classification

An Efficient Method of Hyperspectral Image Dimension Reduction Based on Low Rank Representation and Locally Linear Embedding

Feature Extraction for Hyperspectral Images Using Low-Rank Representation with Neighborhood Preserving Regularization

Tensor Discriminative Locality Alignment for Hyperspectral Image Spectral–Spatial Feature Extraction

Low-Rank Discriminative Embedding Regression for Robust Feature Extraction of Hyperspectral Images via Weighted Schatten p-Norm Minimization

Constrained discriminant neighborhood embedding for high dimensional data feature extraction.

Nonnegative Discriminative Manifold Learning for Hyperspectral Data Dimension Reduction.

Hyperspectral Image Reconstruction of SD-CASSI Based on Nonlocal Low-Rank Tensor Prior

Collaborative Representation Based Local Discriminant Projection for Feature Extraction

Hyperspectral image spectral–spatial classification using local tensor discriminant feature extraction

Unsupervised Feature Extraction for Hyperspectral Images Using Combined Low Rank Representation and Locally Linear Embedding