A two-stage framework combines a beta-conditioned VAE with a latent-space diffusion denoiser to obtain both disentangled representations and high-fidelity image generation.
Since the additive noise for the reparameterization trick follows a normal distribution N (0, Id), the covariance matrix satisfies Σϕ,σ2 = σ2I + Σe ϕ
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Denoising Multi-Beta VAE: Representation Learning for Disentanglement and Generation
A two-stage framework combines a beta-conditioned VAE with a latent-space diffusion denoiser to obtain both disentangled representations and high-fidelity image generation.