Jeremy Warner

EECS Department, University of California, Berkeley

Technical Report No. UCB/EECS-2023-209

August 10, 2023

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-209.pdf

Visual media like vector graphics and presentations are essential to human communication. Artists, designers, and presenters use an iterative reflection and recomposition process to progressively refine their work. Here, reflection describes gathering and absorbing feedback or information after some visual media is created, while recomposition describes adapting and transforming some existing visual media. The key to supporting these stages is providing users with rich, flexible data representations that adhere to their domain and task. Additionally, user interfaces for interacting with this data must balance raw transformative power and the author's creative control. This dissertation supports these media reflection and recomposition processes, and presents techniques for leveraging automatic styling and feedback systems in visual media creation processes.

The goal is to broaden the focus on generative creation tools to the more extensive iterative cycle designers employ (i.e., reflection and recomposition). The first aspect focuses on collecting and refining presentation feedback with SlideSpecs. SlideSpecs is a system for collecting, aggregating, and understanding presentation feedback to assist author refinements. Next, this dissertation will present VST and VLT, a pair of automation-powered design tools for transferring and mixing vector graphics styles and layouts. These tools enable mixing higher-level design properties and rules across vector graphics by generating and leveraging cross-design correspondences. These approaches acknowledge the richness of authors' flexibility and control when shaping their media. With this framing, I demonstrate the benefits and requirements for extending this richness throughout the iterative design cycle.

Advisors: Björn Hartmann

---

BibTeX citation:

@phdthesis{Warner:EECS-2023-209,
    Author= {Warner, Jeremy},
    Title= {Enhancing Visual Media Through Reflection and Recomposition},
    School= {EECS Department, University of California, Berkeley},
    Year= {2023},
    Month= {Aug},
    Url= {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-209.html},
    Number= {UCB/EECS-2023-209},
    Abstract= {Visual media like vector graphics and presentations are essential to human communication. Artists, designers, and presenters use an iterative reflection and recomposition process to progressively refine their work. Here, reflection describes gathering and absorbing feedback or information after some visual media is created, while recomposition describes adapting and transforming some existing visual media. The key to supporting these stages is providing users with rich, flexible data representations that adhere to their domain and task. Additionally, user interfaces for interacting with this data must balance raw transformative power and the author's creative control. This dissertation supports these media reflection and recomposition processes, and presents techniques for leveraging automatic styling and feedback systems in visual media creation processes.

The goal is to broaden the focus on generative creation tools to the more extensive iterative cycle designers employ (i.e., reflection and recomposition). The first aspect focuses on collecting and refining presentation feedback with SlideSpecs. SlideSpecs is a system for collecting, aggregating, and understanding presentation feedback to assist author refinements. Next, this dissertation will present VST and VLT, a pair of automation-powered design tools for transferring and mixing vector graphics styles and layouts. These tools enable mixing higher-level design properties and rules across vector graphics by generating and leveraging cross-design correspondences. These approaches acknowledge the richness of authors' flexibility and control when shaping their media. With this framing, I demonstrate the benefits and requirements for extending this richness throughout the iterative design cycle.},
}

EndNote citation:

%0 Thesis
%A Warner, Jeremy 
%T Enhancing Visual Media Through Reflection and Recomposition
%I EECS Department, University of California, Berkeley
%D 2023
%8 August 10
%@ UCB/EECS-2023-209
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2023/EECS-2023-209.html
%F Warner:EECS-2023-209