Spatially regularized active diffusion learning for high-dimensional images

TL;DR

This paper introduces a spatially-regularized active diffusion learning method for high-dimensional images that efficiently selects influential training points, achieving high accuracy with minimal labels by leveraging diffusion geometry and spatial regularization.

Contribution

It proposes a novel active learning algorithm combining spatial regularization with diffusion geometry to improve label efficiency in high-dimensional image classification.

Findings

Achieves state-of-the-art performance on hyperspectral images.

Produces high-accuracy labels with very few training samples.

Scales linearly with data size under certain models.

Abstract

An active learning algorithm for the classification of high-dimensional images is proposed in which spatially-regularized nonlinear diffusion geometry is used to characterize cluster cores. The proposed method samples from estimated cluster cores in order to generate a small but potent set of training labels which propagate to the remainder of the dataset via the underlying diffusion process. By spatially regularizing the rich, high-dimensional spectral information of the image to efficiently estimate the most significant and influential points in the data, our approach avoids redundancy in the training dataset. This allows it to produce high-accuracy labelings with a very small number of training labels. The proposed algorithm admits an efficient numerical implementation that scales essentially linearly in the number of data points under a suitable data model and enjoys state-of-the-art performance on real hyperspectral images.