D2C: Diffusion-Decoding Models for Few-Shot Conditional Generation

D2C uses a learned diffusion-based prior over the latent representations to improve generation and contrastive self-supervised learning to improve representation quality.