# Learnable Manifold Alignment (LeMA) : A Semi-supervised Cross-modality Learning Framework for Land Cover and Land Use Classification

## TL;DR

LeMA is a semi-supervised framework that learns a joint graph structure from data to improve land cover and land use classification across different remote sensing modalities, especially when labeled hyperspectral data is limited.

## Contribution

The paper introduces LeMA, a novel learnable manifold alignment method that learns data-driven graphs for cross-modality classification, outperforming existing semi-supervised approaches.

## Key findings

- LeMA outperforms state-of-the-art methods on hyperspectral-multispectral datasets.
- Learned graphs better capture data distribution than fixed Gaussian kernels.
- The framework effectively leverages limited hyperspectral data to improve classification accuracy.

## Abstract

In this paper, we aim at tackling a general but interesting cross-modality feature learning question in remote sensing community --- can a limited amount of highly-discrimin-ative (e.g., hyperspectral) training data improve the performance of a classification task using a large amount of poorly-discriminative (e.g., multispectral) data? Traditional semi-supervised manifold alignment methods do not perform sufficiently well for such problems, since the hyperspectral data is very expensive to be largely collected in a trade-off between time and efficiency, compared to the multispectral data. To this end, we propose a novel semi-supervised cross-modality learning framework, called learnable manifold alignment (LeMA). LeMA learns a joint graph structure directly from the data instead of using a given fixed graph defined by a Gaussian kernel function. With the learned graph, we can further capture the data distribution by graph-based label propagation, which enables finding a more accurate decision boundary. Additionally, an optimization strategy based on the alternating direction method of multipliers (ADMM) is designed to solve the proposed model. Extensive experiments on two hyperspectral-multispectral datasets demonstrate the superiority and effectiveness of the proposed method in comparison with several state-of-the-art methods.