Generating High Fidelity Data from Low-density Regions using Diffusion Models

April 05, 2022

Abstract

Our work focuses on addressing sample deficiency from low-density regions of data manifold in common image datasets. We leverage diffusion process based generative models to synthesize novel images from low-density regions. We observe that uniform sampling from diffusion models predominantly samples from high-density regions of the data manifold. Therefore, we modify the sampling process to guide it towards low-density regions while simultaneously maintaining the fidelity of synthetic data. We rigorously demonstrate that our process successfully generates novel high fidelity samples from low-density regions. We further examine generated samples and show that the model does not memorize low-density data and indeed learns to generate novel samples from low-density regions.

Download the Paper

AUTHORS

Written by

Caner Hazirbas

Albert Gordo

Cristian Canton Ferrer

Firat Ozgenel

Vikash Sehwag

Publisher

CVPR

Research Topics

Computer Vision

Core Machine Learning