Exploring Text Rendering Capabilities in the Era of GenAI

PhD Qualifying Examination


Title: "Exploring Text Rendering Capabilities in the Era of GenAI"

by

Mr. Jingye CHEN


Abstract:

In this survey, we investigate the longstanding challenge of generating text 
within im-ages using generative models. Text plays a pivotal role in images, 
with broad applications in posters, covers, advertisements, logos, and more. We 
fully explore existing generative models to understand why they struggle to 
effectively render text, regarding the criteria including accuracy, 
controllability, aesthetic layout, autonomy, text diversity, and layer-wise 
integration. In our study, we divide existing approaches into two categories 
according to the type of generated results: non-layered text image generation 
and layered text image generation. We thoroughly review the designs and 
limitations of these methods. Finally, we draw a conclusion for this survey and 
illustrate promising directions for future research.


Date:                   Friday, 1 November 2024

Time:                   10:00am - 12:00noon

Venue:                  Room 5501
                        Lifts 25/26

Committee Members:      Dr. Qifeng Chen (Supervisor)
                        Prof. Dit-Yan Yeung (Chairperson)
                        Prof. Pedro Sander
                        Prof. Chi-Keung Tang