Logo HyperSIGMA:
Hyperspectral Intelligence Comprehension Foundation Model

Di Wang1,✢, Meiqi Hu1,✢, Yao Jin1,✢, Yuchun Miao1,✢, Jiaqi Yang1,✢, Yichu Xu1,✢, Xiaolei Qin1,✢,
Jiaqi Ma1,✢, Lingyu Sun1,✢, Chenxing Li1,✢, Chuan Fu2, Hongruixuan Chen3,
Chengxi Han1,📧, Naoto Yokoya3, Jing Zhang1,📧, Minqiang Xu4, Lin Liu4,
Lefei Zhang1, Chen Wu1,📧, Bo Du1,📧, Dacheng Tao5, Liangpei Zhang1,📧,

Wuhan University1       Chongqing University2       The University of Tokyo3
National Engineering Research Center of Speech and Language Information Processing4
Nanyang Technological University5
Under Peer Review
✢: Equal Contribution, 📧: Corresponding Author
Code Dataset Model arXiv Media

TL;NR: The first billion-level foundation model
specifically designed for HSI interpretation!

Abstract

The first billion-level foundation model specifically designed for HSI interpretation. To tackle the spectral and spatial redundancy challenges in HSIs, we introduce a novel sparse sampling attention (SSA) mechanism, which effectively promotes the learning of diverse contextual features and serves as the basic block of HyperSIGMA. HyperSIGMA integrates spatial and spectral features using a specially designed spectral enhancement module.

Introduction

Datasets

MY ALT TEXT

We collected hyperspectral remote sensing samples around the globe, constructing a large-scale hyperspectral dataset -- HyperGlobal-450K.
HyperGlobal-450K contains over 20 million three-band images, far exceeding the scale of existing hyperspectral datasets.

Classification

Target Detection

Change Detection

Unmixing

Denoising & Super-resolution

Pairwise Comparisons of Super-resolution Results

Real-world Applicability

MY ALT TEXT

Visualization of detected oil leakage in various Gulf of Mexico regions by our models: (a) GM07, (b) GM13, (c) GM17, and (d) GM18.

BibTeX

@article{hypersigma,
          title={HyperSIGMA: Hyperspectral Intelligence Comprehension Foundation Model},
          author={Wang, Di and Hu, Meiqi and Jin, Yao and Miao, Yuchun and Yang, Jiaqi and Xu, Yichu and Qin, Xiaolei and Ma, Jiaqi and Sun, Lingyu and Li, Chenxing and Fu, Chuan and Chen, Hongruixuan and Han, Chengxi and Yokoya, Naoto and Zhang, Jing and Xu, Minqiang and Liu, Lin and Zhang, Lefei and Wu, Chen and Du, Bo and Tao, Dacheng and Zhang, Liangpei},
          journal={arXiv preprint arXiv:2406.11519},
          year={2024}
      }