PosterMate: Audience-driven Collaborative Persona Agents for Poster Design
summary
PosterMate generates audience-driven persona agents from a marketing brief, lets each agent give contextual text and image edits on a poster canvas, and runs a moderator-facilitated discussion that outputs a single agreed edit. The system is implemented end-to-end for rapid prototyping and image/text preview application; a lab study (N=12) finds it surfaces missed perspectives, and an online study (N=100) shows individual agent feedback matches persona identity above chance and the moderated discussion produces synthesized conclusion preferred over raw suggestions.
AI-written summary · about
chapters (8)
- 0:00Introduction and motivation
- 1:49Problem with traditional audience feedback
- 3:14Formative study findings
- 4:17PosterMate system design
- 6:25User study results
- 7:06Controlled study validation
- 8:30Conclusion and future work
- 9:12Q&A session
related talks
19mBeyond agents: AI as a creative partner
43mHow to Improve your Persuasion Skills with AI Personas and Frameworks
11mBot Personas as Off-The-Shelf Users
01h 33mPoster Pro: Create engaging flipchart posters to wow in workshops
18mPersonas Are Dead. Long Live Personas!
01h 15mPre-sets and Post-sets: Poetics and Projections
09mRethinking AI: Building What Matters, Not Just What's Possible
01h 00mPanel Discussion: AI Meets Design: A Dialogue