Volterra Generative Models Introduce Path-Dependent Noise for Enhanced Diffusion

Researchers have introduced Volterra generative models, a new framework for continuous-time score-based diffusion models. Unlike traditional models that use Brownian perturbations, Volterra models incorporate path-dependent noise through fractional kernels. To manage these non-Markovian dynamics, the framework employs finite-dimensional Markovian lifts and Gaussian quadrature. Experiments on MNIST and CIFAR-10 datasets indicate that these persistent fractional perturbations can enhance score-based generation, with a bridge sampler offering stability for larger lifts. AI

IMPACT Introduces a novel approach to diffusion models with potential for improved image generation.

RANK_REASON The cluster contains an academic paper detailing a new generative model framework. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• CIFAR-10
• MNIST
• Volterra generative models

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →