Semi-Supervised Superpixel-Based Multi-Feature Graph Learning for Hyperspectral Image Data

Graphs naturally lend themselves to model the complexities of Hyperspectral Image (HSI) data as well serve semi-supervised classifiers by propagating given labels among nearest neighbours. In this work, we present a novel framework for classification HSI in light very limited amount labelled data, inspired multi-view graph learning and signal processing. Given an priori superpixel-segmented hyperspectral image, seek robust efficient construction label propagation method conduct (SSL). Since is paramount success subsequent task, particularly intrinsic complexity consider problem finding optimal such data. Our contribution two-fold: firstly, propose multi-stage edge-efficient which exploits information through pseudo-label features embedded construction. Secondly, examine enhance multiple superpixel on basis pseudo-labels extension previous framework, less reliant excessive parameter tuning. Ultimately, demonstrate superiority our approaches comparison with state-of-the-art methods extensive numerical experiments.