Empirical study on CLD baseline shows purely synthetic SynLayers data outperforms PrismLayersPro for layer decomposition, with gains scaling to ~50K samples and better layer-count balance.
For each base, we retain the background canvas and randomly remove Nremove ∼ U {1,4} foreground layers while ensuring at least one remains
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Does Synthetic Layered Design Data Benefit Layered Design Decomposition?
Empirical study on CLD baseline shows purely synthetic SynLayers data outperforms PrismLayersPro for layer decomposition, with gains scaling to ~50K samples and better layer-count balance.