Can We Change the Stroke Size for Easier Diffusion?

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bai, Yunwei, Tan, Ying Kiat, Shu, Yao, Chen, Tsuhan
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910171758854144
author Bai, Yunwei
Tan, Ying Kiat
Shu, Yao
Chen, Tsuhan
author_facet Bai, Yunwei
Tan, Ying Kiat
Shu, Yao
Chen, Tsuhan
contents Diffusion models can be challenged in the low signal-to-noise regime, where they have to make pixel-level predictions despite the presence of high noise. The geometric intuition is akin to using the finest stroke for oil painting throughout, which may be ineffective. We therefore study stroke-size control as a controlled intervention that changes the effective roughness of the supervised target, predictions and perturbations across timesteps, in an attempt to ease the low signal-to-noise challenge.
format Preprint
id arxiv_https___arxiv_org_abs_2603_26783
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Can We Change the Stroke Size for Easier Diffusion?
Bai, Yunwei
Tan, Ying Kiat
Shu, Yao
Chen, Tsuhan
Computer Vision and Pattern Recognition
Artificial Intelligence
Diffusion models can be challenged in the low signal-to-noise regime, where they have to make pixel-level predictions despite the presence of high noise. The geometric intuition is akin to using the finest stroke for oil painting throughout, which may be ineffective. We therefore study stroke-size control as a controlled intervention that changes the effective roughness of the supervised target, predictions and perturbations across timesteps, in an attempt to ease the low signal-to-noise challenge.
title Can We Change the Stroke Size for Easier Diffusion?
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2603.26783