Salvato in:
Dettagli Bibliografici
Autori principali: Grimal, Paul, Borgne, Hervé Le, Ferret, Olivier
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2504.17525
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912344268865536
author Grimal, Paul
Borgne, Hervé Le
Ferret, Olivier
author_facet Grimal, Paul
Borgne, Hervé Le
Ferret, Olivier
contents Visual generative AI models often encounter challenges related to text-image alignment and reasoning limitations. This paper presents a novel method for selectively enhancing the signal at critical denoising steps, optimizing image generation based on input semantics. Our approach addresses the shortcomings of early-stage signal modifications, demonstrating that adjustments made at later stages yield superior results. We conduct extensive experiments to validate the effectiveness of our method in producing semantically aligned images on Diffusion and Flow Matching model, achieving state-of-the-art performance. Our results highlight the importance of a judicious choice of sampling stage to improve performance and overall image alignment.
format Preprint
id arxiv_https___arxiv_org_abs_2504_17525
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Text-to-Image Alignment in Denoising-Based Models through Step Selection
Grimal, Paul
Borgne, Hervé Le
Ferret, Olivier
Computer Vision and Pattern Recognition
Visual generative AI models often encounter challenges related to text-image alignment and reasoning limitations. This paper presents a novel method for selectively enhancing the signal at critical denoising steps, optimizing image generation based on input semantics. Our approach addresses the shortcomings of early-stage signal modifications, demonstrating that adjustments made at later stages yield superior results. We conduct extensive experiments to validate the effectiveness of our method in producing semantically aligned images on Diffusion and Flow Matching model, achieving state-of-the-art performance. Our results highlight the importance of a judicious choice of sampling stage to improve performance and overall image alignment.
title Text-to-Image Alignment in Denoising-Based Models through Step Selection
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.17525