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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Learning effective spectral-spatial features is important for thehyperspectral image (HSI) classification task, but the majority of existing HSIclassification methods still suffer from modeling complex spectral-spatialrelations and characterizing low-level details and high-level semanticscomprehensively. As a new class of record-breaking generative models, diffusionmodels are capable of modeling complex relations for understanding inputs wellas learning both high-level and low-level visual features. Meanwhile, diffusionmodels can capture more abundant features by taking advantage of the extra andunique dimension of timestep t. In view of these, we propose an unsupervisedspectral-spatial feature learning framework based on the diffusion model forHSI classification for the first time, named Diff-HSI. Specifically, we firstpretrain the diffusion model with unlabeled HSI patches for unsupervisedfeature learning, and then exploit intermediate hierarchical features fromdifferent timesteps for classification. For better using the abundanttimestep-wise features, we design a timestep-wise feature bank and a dynamicfeature fusion module to construct timestep-wise features, adaptively learninginformative multi-timestep representations. Finally, an ensemble of linearclassifiers is applied to perform HSI classification. Extensive experiments areconducted on three public HSI datasets, and our results demonstrate thatDiff-HSI outperforms state-of-the-art supervised and unsupervised methods forHSI classification.

When Hyperspectral Image Classification Meets Diffusion Models: An Unsupervised Feature Learning Framework