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

03/31/2022
by   Vikash Sehwag, et al.
1

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.

READ FULL TEXT

page 1

page 6

page 12

page 14

page 15

page 16

page 17

page 18

research
01/29/2023

Don't Play Favorites: Minority Guidance for Diffusion Models

We explore the problem of generating minority samples using diffusion mo...
research
01/31/2023

Generating High Fidelity Synthetic Data via Coreset selection and Entropic Regularization

Generative models have the ability to synthesize data points drawn from ...
research
10/23/2022

Deep Equilibrium Approaches to Diffusion Models

Diffusion-based generative models are extremely effective in generating ...
research
02/14/2018

Geometry-Based Data Generation

Many generative models attempt to replicate the density of their input d...
research
06/12/2023

Latent Dynamical Implicit Diffusion Processes

Latent dynamical models are commonly used to learn the distribution of a...
research
07/30/2020

Instance Selection for GANs

Recent advances in Generative Adversarial Networks (GANs) have led to th...
research
01/16/2023

Msanii: High Fidelity Music Synthesis on a Shoestring Budget

In this paper, we present Msanii, a novel diffusion-based model for synt...

Please sign up or login with your details

Forgot password? Click here to reset