A spatial-spectral transformer that uses finite differences of attention scores between neighboring patches for hyperspectral image classification.
An overview on spectral and spatial information fusion for hyperspectral image classification: Current trends and challenges,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
DiffFormer: a Differential Spatial-Spectral Transformer for Hyperspectral Image Classification
A spatial-spectral transformer that uses finite differences of attention scores between neighboring patches for hyperspectral image classification.