Analysis of Classifier-Free Guidance Weight Schedulers

Classifier-Free Guidance (CFG) enhances the quality and condition adherence of text-toimage diffusion models. It operates by combining the conditional and unconditional predictions using a fixed weight. However, recent works vary the weights throughout the diffusion process, reporting superior resul...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Transactions on Machine Learning Research Journal
Hauptverfasser: Wang, Xi, Dufour, Nicolas, Andreou, Nefeli, Cani, Marie-Paule, Fernández Abrevaya, Victoria, Picard, David, Kalogeiton, Vicky
Format: Journal Article
Sprache:Englisch
Veröffentlicht: [Amherst Massachusetts]: OpenReview.net, 2022 2024
Schlagworte:
Artificial Intelligence
Computer Science
Generative Model
Diffusion Model
ISSN:2835-8856
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Classifier-Free Guidance (CFG) enhances the quality and condition adherence of text-toimage diffusion models. It operates by combining the conditional and unconditional predictions using a fixed weight. However, recent works vary the weights throughout the diffusion process, reporting superior results but without providing any rationale or analysis. By conducting comprehensive experiments, this paper provides insights into CFG weight schedulers. Our findings suggest that simple, monotonically increasing weight schedulers consistently lead to improved performances, requiring merely a single line of code. In addition, more complex parametrized schedulers can be optimized for further improvement, but do not generalize across different models and tasks.
ISSN:2835-8856
DOI:10.48550/arXiv.2404.13040