Domain Adaptor Networks for Hyperspectral Image Recognition
This work addresses the challenge of hyperspectral image recognition for researchers and practitioners, offering incremental advancements in domain adaptation techniques.
The paper tackles the problem of adapting networks trained on color images to hyperspectral domains with many channels, proposing domain adaptor networks that enable learning on small datasets and finding that simple schemes like linear projection are often effective but can sometimes reduce performance, with a novel multi-view adaptor providing further improvements.
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.