Domain Adaptor Networks for Hyperspectral Image Recognition

TL;DR

This paper introduces domain adaptor networks that enable the adaptation of color image recognition models to hyperspectral images, improving performance on small datasets through novel architectures and strategies.

Contribution

It proposes new domain adaptor architectures, including a multi-view adaptor, for hyperspectral image recognition, and provides extensive experimental evaluation.

Findings

Simple adaptors like linear projection are effective but may reduce accuracy.

Multi-view adaptors improve recognition performance.

Trade-offs between accuracy and computational cost are characterized.

Abstract

We consider the problem of adapting a network trained on three-channel color images to a hyperspectral domain with a large number of channels. To this end, we propose domain adaptor networks that map the input to be compatible with a network trained on large-scale color image datasets such as ImageNet. Adaptors enable learning on small hyperspectral datasets where training a network from scratch may not be effective. We investigate architectures and strategies for training adaptors and evaluate them on a benchmark consisting of multiple hyperspectral datasets. We find that simple schemes such as linear projection or subset selection are often the most effective, but can lead to a loss in performance in some cases. We also propose a novel multi-view adaptor where of the inputs are combined in an intermediate layer of the network in an order invariant manner that provides further improvements. We present extensive experiments by varying the number of training examples in the benchmark to characterize the accuracy and computational trade-offs offered by these adaptors.