HyperKD transfers knowledge from a multispectral foundation model to a hyperspectral masked autoencoder, improving reconstruction and several downstream tasks over a non-distilled baseline.
The enmap imaging spectroscopy mission towards operations,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
HyperKD: Distilling Cross-Spectral Knowledge in Masked Autoencoders via Inverse Domain Shift with Spatial-Aware Masking and Specialized Loss
HyperKD transfers knowledge from a multispectral foundation model to a hyperspectral masked autoencoder, improving reconstruction and several downstream tasks over a non-distilled baseline.