# Transfer Learning for Segmenting Dimensionally-Reduced Hyperspectral Images

## TL;DR

This paper presents a transfer learning approach combined with spectral dimensionality reduction to effectively train deep neural networks for hyperspectral image segmentation using limited ground-truth data across different sensors.

## Contribution

It introduces a novel method that leverages transfer learning and spectral reduction to improve hyperspectral segmentation with scarce data and sensor variability.

## Key findings

- Effective training of deep nets with limited data
- Robust segmentation across different sensors
- Statistically validated performance improvements

## Abstract

Deep learning has established the state of the art in multiple fields, including hyperspectral image analysis. However, training large-capacity learners to segment such imagery requires representative training sets. Acquiring such data is human-dependent and time-consuming, especially in Earth observation scenarios, where the hyperspectral data transfer is very costly and time-constrained. In this letter, we show how to effectively deal with a limited number and size of available hyperspectral ground-truth sets, and apply transfer learning for building deep feature extractors. Also, we exploit spectral dimensionality reduction to make our technique applicable over hyperspectral data acquired using different sensors, which may capture different numbers of hyperspectral bands. The experiments, performed over several benchmarks and backed up with statistical tests, indicated that our approach allows us to effectively train well-generalizing deep convolutional neural nets even using significantly reduced data.