A combined DenseNet-residual-transformer encoder-decoder model produces competitive precipitation nowcasts over North-East India with 3.3M parameters.
Here, we employed the transformer-based attention mechanism to leverage the efficiency of a CNN -based encoder-decoder model
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Dense-Cast: A lightweight ensemble of deep learning architectures for precipitation nowcasting
A combined DenseNet-residual-transformer encoder-decoder model produces competitive precipitation nowcasts over North-East India with 3.3M parameters.