A scalar bias added to the latent noise of a diffusion model, tuned per target dataset, improves counterfactual medical image generation after fine-tuning.
Sutherland, Michael Arbel, and Arthur Gretton
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
-
Latent Drifting in Diffusion Models for Counterfactual Medical Image Synthesis
A scalar bias added to the latent noise of a diffusion model, tuned per target dataset, improves counterfactual medical image generation after fine-tuning.