Students from Associate Professor Nao Tokui's Lab at the Graduate School of Media and Governance Selected for Artwork Spotlights at "Machine Learning for Creativity and Design"

January 31, 2023

In the art track of the "Machine Learning for Creativity and Design" workshop, hosted by the international AI conference NeurIPS 2022, artworks by students from Nao Tokui's lab were selected as Artwork Spotlights.

The "Machine Learning for Creativity and Design" workshop was held online in December 2022.

This year, five works were selected as Artwork Spotlights from a total of 43 submissions, and two of them were by students from the Tokui Lab.

Keito Takaishi (2nd-year master's student, Graduate School of Media and Governance)

Kazufumi Shibuya (3rd-year student, Faculty of Environment and Information Studies)

Asuka Ishii (3rd-year student, Faculty of Policy Management)

We are honored to have been selected for the Artwork Spotlights.

Recently, AI that generates high-definition images from text has become widely used. In such services, the generation of sexual or grotesque images is prohibited, and words related to these topics are blocked. However, just as eroticism, a subject of regulation, has been a theme in art since ancient times, the restricted expressive domain of AI must contain not only the obscene and ugly but also things that are aesthetic while defying conventional morality. This work is a 3D piece created by inputting text containing prohibited words into an AI model that constructs 3D data from text, then generating and printing the result. By using sexual words in parts of the text, we generated a physical curvilinear beauty on parts of the 3D data's surface, pursuing the visual beauty hidden within the expressive domain restricted by the prohibition of certain words. We also attempted to salvage this beauty existing within the AI into the real world in an uneditable state through 3D printing.

We would like to express our deepest gratitude to Professor Tokui for his guidance on this project, and to everyone else who supported us. (Keito Takaishi)

Yota Naruse (4th-year student, Faculty of Environment and Information Studies)

I am very honored to have been selected for the Artwork Spotlights at the "Machine Learning for Creativity and Design" workshop. I believe this work was recognized for its focus on the interaction between artificial intelligence and humans, and for constructing a structure of co-creation between them by combining it with audio-visual expression. I will continue to devote myself to my creative work and research.

I would like to thank everyone in the lab for their guidance on this project. Thank you very much.

Machine Learning for Creativity and Design

Nao Tokui's Lab

Nao Tokui, Associate Professor, Graduate School of Media and Governance Profile

Source: General Affairs Section, Shonan Fujisawa Campus (SFC) Office