English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This work studies Hyperspectral image (HSI) super-resolution (SR). HSI SR ischaracterized by high-dimensional data and a limited amount of trainingexamples. This exacerbates the undesirable behaviors of neural networks such asmemorization and sensitivity to out-of-distribution samples. This workaddresses these issues with three contributions. First, we observe that HSI SRand RGB image SR are correlated and develop a novel multi-tasking network totrain them jointly so that the auxiliary task RGB image SR can provideadditional supervision. Second, we propose a simple, yet effective dataaugmentation routine, termed Spectral Mixup, to construct effective virtualtraining samples to enlarge the training set. Finally, we extend the network toa semi-supervised setting so that it can learn from datasets containing onlylow-resolution HSIs. With these contributions, our method is able to learn fromheterogeneous datasets and lift the requirement for having a large amount of HDHSI training samples. Extensive experiments on four standard datasets show thatour method outperforms existing methods significantly and underpin therelevance of our contributions. Code has been made available athttps://github.com/kli8996/HSISR.

Hyperspectral Image Super-Resolution with Spectral Mixup and Heterogeneous Datasets