Hyperspectral Image Classification Based on Adaptive Sparse Deep Network

TL;DR

This paper introduces an adaptive sparse deep network for hyperspectral image classification that automatically optimizes sparse representation and regularization parameters, leading to improved classification accuracy.

Contribution

It proposes a novel deep architecture that constructs optimal sparse representations and regularization parameters using a data flow graph and ADMM, with parameters updated via back-propagation.

Findings

Achieves significant improvement over traditional classifiers.

Effectively constructs optimal sparse representations.

Demonstrates superior classification performance on hyperspectral images.

Abstract

Sparse model is widely used in hyperspectral image classification.However, different of sparsity and regularization parameters has great influence on the classification results.In this paper, a novel adaptive sparse deep network based on deep architecture is proposed, which can construct the optimal sparse representation and regularization parameters by deep network.Firstly, a data flow graph is designed to represent each update iteration based on Alternating Direction Method of Multipliers (ADMM) algorithm.Forward network and Back-Propagation network are deduced.All parameters are updated by gradient descent in Back-Propagation.Then we proposed an Adaptive Sparse Deep Network.Comparing with several traditional classifiers or other algorithm for sparse model, experiment results indicate that our method achieves great improvement in HSI classification.