Synthetic images from diffusion models, used as pre-training data, substantially improve polyp localization when real data is scarce, with the benefit shrinking as more real images are added.
A multi-centre polyp detection and segmentation dataset for generalisability assessment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Data Augmentation with Diffusion Models for Colon Polyp Localization on the Low Data Regime: How much real data is enough?
Synthetic images from diffusion models, used as pre-training data, substantially improve polyp localization when real data is scarce, with the benefit shrinking as more real images are added.