Image Interpolation with Score-based Riemannian Metrics of Diffusion Models
Track: tiny / short paper (up to 4 pages)
Keywords: pre-trained diffusion models, Riemannian geometry, manifold hypothesis, image interpolation
TL;DR: Smooth image interpolation by introducing a Riemannian metric in data space derived from the score function of a pre-trained diffusion model.
Abstract: Diffusion models excel in content generation by implicitly learning the data manifold, yet they lack a practical method to leverage this manifold---unlike other deep generative models equipped with latent spaces. This paper introduces a novel framework that treats the data space of pre-trained diffusion models as a Riemannian manifold, with a metric derived from score function. Experiments with MNIST and Stable Diffusion show that this geometry-aware approach yields smoother interpolations than linear or spherical linear interpolation and other methods, demonstrating its potential for improved content generation and editing.
Submission Number: 118
Loading