WHU-Sigma/HyperSIGMA
The official repo for [TPAMI'25] "HyperSIGMA: Hyperspectral Intelligence Comprehension Foundation Model"
Combines a billion-parameter vision transformer with sparse sampling attention (SSA) to reduce spectral-spatial redundancy in hyperspectral imagery, enabling efficient feature learning across diverse bands. Integrates a spectral enhancement module for joint spatial-spectral representation and supports both high-level tasks (classification, segmentation) and low-level restoration (denoising, super-resolution). Pre-trained on HyperGlobal-450K dataset and distributed via HuggingFace and Baidu Drive with checkpoints for downstream HSI and multispectral applications.
346 stars.
Stars
346
Forks
28
Language
Python
License
Apache-2.0
Category
Last pushed
Jan 18, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/WHU-Sigma/HyperSIGMA"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
manideep2510/eye-in-the-sky
Satellite Image Classification using semantic segmentation methods in deep learning
rogerxujiang/dstl_unet
Dstl Satellite Imagery Feature Detection
motokimura/spacenet_building_detection
Project to train/test convolutional neural networks to extract buildings from SpaceNet satellite...
Paulymorphous/skeyenet
Road and Building Segmentation in Satellite Imagery
IGNF/FLAIR-1
Semantic segmentation from aerial imagery (baseline of the FLAIR #1 challenge)