Personalizing Products with Stylized Head Portraits for Self-Expression

要旨

Personalizing products aesthetically or functionally can help users increase personal relevance and support self-expression. However, using non-abstract personal data such as head portraits for product personalization has been understudied. While recent advances in Artificial Intelligence have enabled generating stylized head portraits, these images also raise concerns about lack of control, artificiality, and ethics, which potentially limit their broader use. In this work, we present PicMe, a design support tool that converts user face photos into stylized head portraits as vector graphics that can be used to personalize products. To enable style transfer, PicMe leverages a deep-learning-based algorithm trained on an extended open-source illustration dataset of characters in a cartoonish and minimalistic style. We evaluated PicMe through two experiments and a user study. The results of our evaluation showed that PicMe can help create personalized head portraits that support self-expression.

著者
Yang Shi
College of Design and Innovation, Tongji University, Shanghai, China
Yechun Peng
Tongji University, Shanghai, China
Shengqi Dang
College of Design and Innovation, Tongji University, Shanghai, China
Nanxuan Zhao
Harvard University, Cambridge, Massachusetts, United States
Nan Cao
Tongji College of Design and Innovation, Shanghai, China
論文URL

https://doi.org/10.1145/3613904.3642391

動画

会議: CHI 2024

The ACM CHI Conference on Human Factors in Computing Systems (https://chi2024.acm.org/)

セッション: Creativity Tools

323C
5 件の発表
2024-05-16 01:00:00
2024-05-16 02:20:00