LogoMotion vs. The Newbies (Spoiler: The Newbies Won)
This section evaluates the usability of the LogoMotion system by novices in the field of animation. The primary research question (RQ5) is whether user interaction and iteration can enhance the quality of automatically generated logos. Specifically, the…
This section evaluates the usability of the LogoMotion system by novices in the field of animation. The primary research question (RQ5) is whether user interaction and iteration can enhance the quality of automatically generated logos. Specifically, the study examines the usability of the system for novices and the number of iterations required to achieve satisfactory design outcomes.
The study recruited five animation novices from a university. Participants selected two templates from a gallery to begin their tasks. They assessed the results generated by LogoMotion, identifying which were publishable without changes and which required further edits. Participants were able to input prompts into a textbox, initiating edits to the animation code.
The findings indicate that many novices were satisfied with the system-generated animations. Out of 16 animations, 10 were deemed usable without modification. Participants appreciated the ability to customize designs quickly and iterate on the initial concepts. For example, changing the entry angle of a skier in an animation or adjusting the timing of elements provided valuable creative control.
Novices favored the option to make edits using natural language, impacting various dimensions such as visual, grouping, and ordering changes. Examples of user inputs included instructions like, “have a slower delay at the beginning and don’t move the bird at the end," and “make the text show quicker and don’t have a pulse animation at the end."
This section evaluates the usability of the LogoMotion system by novices in the field of animation.
The system provided four variations for each logo, which allowed users to explore different layouts and directions. Participants found inspiration from the gallery options and often combined elements from multiple animations to create new designs.
Some challenges were noted, particularly in achieving complex interactions between objects and encountering visual bugs during edits. Despite these challenges, all participants were satisfied with at least one animated logo after one or two iterations.
Vivian Liu, Columbia University Rubaiat Habib Kazi, Adobe Research Li-Yi Wei, Adobe Research Matthew Fisher, Adobe Research Timothy Langlois, Adobe Research Seth Walker, Adobe Research Lydia Chilton, Columbia University
The full paper is available on arXiv under the CC BY-NC-ND 4.0 DEED license.
Based on reporting by hackernoon.com.
