CL-IMS @ DIACR-Ita: Volente o Nolente: BERT does not outperform SGNS on Semantic Change Detection.
Severin LaicherGioia BaldissinEnrique CastañedaDominik SchlechtwegSabine Schulte im WaldePublished in: CoRR (2020)
Keyphrases
- change detection
- remote sensing
- remote sensing images
- satellite images
- data streams
- remote sensing imagery
- remotely sensed images
- satellite imagery
- land cover change
- remotely sensed
- land cover
- semantic similarity
- image registration
- multiresolution
- feature extraction
- machine learning
- remotely sensed data
- gamma distributions
- shot change detection
- database