Don't be so negative! Score-based Generative Modeling with Oracle-assisted Guidance

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Naderiparizi, Saeid, Liang, Xiaoxuan, Cohan, Setareh, Zwartsenberg, Berend, Wood, Frank
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912477434871808
author Naderiparizi, Saeid
Liang, Xiaoxuan
Cohan, Setareh
Zwartsenberg, Berend
Wood, Frank
author_facet Naderiparizi, Saeid
Liang, Xiaoxuan
Cohan, Setareh
Zwartsenberg, Berend
Wood, Frank
contents Score-based diffusion models are a powerful class of generative models, widely utilized across diverse domains. Despite significant advancements in large-scale tasks such as text-to-image generation, their application to constrained domains has received considerably less attention. This work addresses model learning in a setting where, in addition to the training dataset, there further exists side-information in the form of an oracle that can label samples as being outside the support of the true data generating distribution. Specifically we develop a new denoising diffusion probabilistic modeling methodology, Gen-neG, that leverages this additional side-information. Gen-neG builds on classifier guidance in diffusion models to guide the generation process towards the positive support region indicated by the oracle. We empirically establish the utility of Gen-neG in applications including collision avoidance in self-driving simulators and safety-guarded human motion generation.
format Preprint
id arxiv_https___arxiv_org_abs_2307_16463
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Don't be so negative! Score-based Generative Modeling with Oracle-assisted Guidance
Naderiparizi, Saeid
Liang, Xiaoxuan
Cohan, Setareh
Zwartsenberg, Berend
Wood, Frank
Machine Learning
Score-based diffusion models are a powerful class of generative models, widely utilized across diverse domains. Despite significant advancements in large-scale tasks such as text-to-image generation, their application to constrained domains has received considerably less attention. This work addresses model learning in a setting where, in addition to the training dataset, there further exists side-information in the form of an oracle that can label samples as being outside the support of the true data generating distribution. Specifically we develop a new denoising diffusion probabilistic modeling methodology, Gen-neG, that leverages this additional side-information. Gen-neG builds on classifier guidance in diffusion models to guide the generation process towards the positive support region indicated by the oracle. We empirically establish the utility of Gen-neG in applications including collision avoidance in self-driving simulators and safety-guarded human motion generation.
title Don't be so negative! Score-based Generative Modeling with Oracle-assisted Guidance
topic Machine Learning
url https://arxiv.org/abs/2307.16463