AI World Museum

AI World Museum, 2021

Type

AI Research, Interactive Game

Key Responsibilities:

· AI Research · Social Studies Research

· Software Development

How many art paintings does a human sees in their lifetime? Would an organised collection of art paintings combined with personalised experience increase the interest in arts? Currently there are few online databases of art paintings and none of them offer a personalised experience. This research project combines Engineering and Social Sciences studies to create a prototype application-experience with an organised art collection.

This research suggests a time-measurement Preference system which is non disturbing to the viewers. Viewers who enter the experience are called to answer a questionnaire and are assigned to a cluster, which will be updated with their preferences upon finishing their virtual visit.

As part of the “Decision Making and Learning in Multi-Agent Worlds” course at TU Crete for the M.Sc. by Research program, I engaged in a semester research project under the guidance of Professor Georgios Chalkiadakis. The project, tailored to individual interests within the course, explored the dynamic intersection of Artificial Intelligence in decision-making and learning in multi-agent environments.

Nestled in the heart of an isolated forest, a cutting-edge 3D museum emerges as a beacon of artistry. Composed mostly of glass, the ethereal structure is home to an ever-evolving collection curated based on the unique preferences of User Categories (UC), with artworks strategically placed using a dynamic list of PlaceHolders. The system meticulously tracks visitor interactions, translating time spent at each painting into weighted values through custom algorithms, seamlessly updating an online Firebase database for real-time insights. As the museum thrives on user engagement, a dynamic Lambda function adjusts preferences based on visitor counts within each UC, with added weight for seasoned enthusiasts to elevate the impact of their insights. This immersive blend of art, technology, and nature promises a personalized experience that evolves with each visitor’s nuanced exploration.

In this haven of creativity, the marriage of data and aesthetics invites all to witness the Art Movements (AM) tailored to their UC. The museum, initially fueled by simulated data, dynamically allocates preferences across various AMs. This ensures a captivating journey for visitors, as the Lambda function adapts to changing dynamics, amplifying the influence of expert opinions. As the transparent confines of the museum continue to draw individuals into its ever-shifting narrative, it stands as a testament to the seamless integration of creativity and technology in an enchanting wilderness setting.

This research project has been a valuable avenue for exploring diverse subjects, spanning engineering to social sciences. The integration of Social Sciences has proven crucial, shaping the project’s trajectory by contributing insightful questions for classification and introducing time measurement as an innovative alternative to traditional rating systems.

The adaptable Preference system showcased in the project holds promise for broader applications. With minimal adjustments, the system could pivot to use narrations as input material, enabling viewers to engage in exploratory journeys through dynamically generated stories, showcasing its versatility beyond the realm of visual art.

Research Advisor: Prof. Georgios Chalkiadakis
AI Research: Yannis Kritikos
Software Development: Yannis Kritikos
3D Building: Savvas Sammanidis