CAIM-Net predicts the change moment from satellite image time series and then derives the change area as the set of pixels with a detected moment, reporting small Kappa gains on DynamicEarthNet and SpaceNet7.
We employ the Adaptive moment estimation (Adam) optimizer (Kingma and Ba, 2017) with an initial learning rate of 1๐๐โ4
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Information transmission: Inferring change area from change moment in time series remote sensing images
CAIM-Net predicts the change moment from satellite image time series and then derives the change area as the set of pixels with a detected moment, reporting small Kappa gains on DynamicEarthNet and SpaceNet7.