On a one-year Stanford sky-image dataset, a frozen pretrained vision transformer predicts solar PV output worse than the CNN baseline (overall RMSE 3.35 vs 2.36), contradicting the paper's own 'almost as well' framing.
Images are then down- sampled to a desired size (64 × 64, 128 × 128, 256 × 256, etc.)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CE 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Transformers Applied to Short-term Solar PV Power Output Forecasting
On a one-year Stanford sky-image dataset, a frozen pretrained vision transformer predicts solar PV output worse than the CNN baseline (overall RMSE 3.35 vs 2.36), contradicting the paper's own 'almost as well' framing.