We introduce Score identity Distillation (SiD), an innovative data-free method that distills the
generative capabilities of pretrained diffusion models into a single-step generator. SiD not …

Score identity Distillation (SiD) is a data-free method that has achieved SOTA performance
in image generation by leveraging only a pretrained diffusion model, without requiring any …

Diffusion-based text-to-image generation models trained on extensive text-image pairs have
shown the capacity to generate photorealistic images consistent with textual descriptions …

While conditional diffusion models have achieved remarkable success in various
applications, they require abundant data to train from scratch, which is often infeasible in …

Variational autoencoders (VAEs) are popular likelihood-based generative models which can
be efficiently trained by maximising an evidence lower bound. There has been much …

Semi-implicit variational inference (SIVI) extends traditional variational families with semi-
implicit distributions defined in a hierarchical manner. Due to the intractable densities of …