Ambient Diffusion Omni: Training Good Models with Bad Data

NeurIPSSpotlight2025

Authors: Giannis Daras, Adrian Rodriguez-Munoz, Adam Klivans, Antonio Torralba, Constantinos Costis Daskalakis
Venue: NeurIPS 2025
Track: Spotlight

TL;DR

We show how to use low-quality, synthetic, and out-of-distribution images to improve the quality of a diffusion model…

Opening excerpt from the authors’ abstract. source

Read the paper

OpenReview

Topics

diffusion

← All NeurIPS 2025 Spotlight papers · Browse the whole archive