ILDM performs diffusion that switches between curved-manifold and flat-space dynamics in a learned latent space, improving FID/LPIPS on small image datasets.
A connection between score matching and denoising autoencoders,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
other 1
citation-polarity summary
fields
stat.ML 1years
2026 1verdicts
CONDITIONAL 1roles
other 1polarities
unclear 1representative citing papers
citing papers explorer
-
Intrinsic-Hybrid Latent Diffusion Models for Generative Modeling on Unknown Manifolds
ILDM performs diffusion that switches between curved-manifold and flat-space dynamics in a learned latent space, improving FID/LPIPS on small image datasets.