Trans-Dimensional Generative Modeling via Jump Diffusion Models

05/25/2023
by   Andrew Campbell, et al.
0

We propose a new class of generative models that naturally handle data of varying dimensionality by jointly modeling the state and dimension of each datapoint. The generative process is formulated as a jump diffusion process that makes jumps between different dimensional spaces. We first define a dimension destroying forward noising process, before deriving the dimension creating time-reversed generative process along with a novel evidence lower bound training objective for learning to approximate it. Simulating our learned approximation to the time-reversed generative process then provides an effective way of sampling data of varying dimensionality by jointly generating state values and dimensions. We demonstrate our approach on molecular and video datasets of varying dimensionality, reporting better compatibility with test-time diffusion guidance imputation tasks and improved interpolation capabilities versus fixed dimensional models that generate state values and dimensions separately.

READ FULL TEXT

page 7

page 9

page 32

page 34

page 35

page 37

research
02/20/2023

Infinite-Dimensional Diffusion Models for Function Spaces

We define diffusion-based generative models in infinite dimensions, and ...
research
11/29/2022

Dimensionality-Varying Diffusion Process

Diffusion models, which learn to reverse a signal destruction process to...
research
07/04/2023

ProtoDiffusion: Classifier-Free Diffusion Guidance with Prototype Learning

Diffusion models are generative models that have shown significant advan...
research
12/01/2022

Diffusion Generative Models in Infinite Dimensions

Diffusion generative models have recently been applied to domains where ...
research
01/27/2023

Minimizing Trajectory Curvature of ODE-based Generative Models

Recent ODE/SDE-based generative models, such as diffusion models, rectif...
research
09/28/2022

Spectral Diffusion Processes

Score-based generative modelling (SGM) has proven to be a very effective...
research
06/01/2021

Hybrid Generative Models for Two-Dimensional Datasets

Two-dimensional array-based datasets are pervasive in a variety of domai...

Please sign up or login with your details

Forgot password? Click here to reset