Triadic Dynamics Aware Diffusion Posterior Sampling for Inverse Problems: Optimizing Guidance and Stochasticity Schedules

Junseo Bang, Dong Ju Mun, Hoigi Seo, Seongmin Hong, Se Young Chun · 2026

Read Triadic Dynamics Aware Diffusion Posterior Sampling for Inverse Problems: Optimizing Guidance and Stochasticity Schedules by Junseo Bang, Dong Ju Mun, Hoigi Seo, Seongmin Hong, Se Young Chun (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-26470.txt.

Description

Junseo Bang and four colleagues schedule diffusion posterior sampling for imaging inverse problems, not by habit. Data-consistency guidance (DC), classifier-free guidance (CFG), and stochasticity interact: strong CFG early clashes with DC, while stochasticity returns the path to higher-probability regions. Their TriPS method recasts sampling as a shifting control problem, following falling DC and noise scales beside rising CFG. Template search gives sound baseline schedules, and group-relative policy optimization learns freer temporal curves. Tests beat baselines on fidelity and perceptual realism.

Catalog record

CREATORJunseo Bang, Dong Ju Mun, Hoigi Seo, Seongmin Hong, Se Young Chun
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE90195
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.26470

Browse the full library · About the Tower of Babel