AI museum tour guide
AI & Data
Semester programme:Minor AI for Society
Client company:Philips Museum
Tomas Hertogh
Kelvin Kanen
Eric van der Wiel
Haakon Valland
Project description
Our project is an interactive AI-powered web application designed for use in the Philips Museum. Visitors can use their smartphone to scan exhibition objects, which are recognized in real time using a custom-trained object detection model. Once an object is identified, users can interact with a chatbot to learn more about its history, function, and relevance. Making museum visits more engaging and informative.
Context
The project was developed in collaboration with the Philips Museum to enhance visitor engagement through AI technology. It operates within the domain of interactive museum experiences, combining computer vision and conversational AI to make cultural heritage more accessible and informative for a broad audience.
Results
The most important outcomes of the project are:
- AI-Powered Web Application: A mobile-friendly app that recognizes museum objects and provides interactive information.
- Custom Object Detection Model: A YOLOv8 model trained to identify over 30 Philips exhibition items in real-time.
- Chatbot Integration: A conversational assistant that answers visitor questions about detected objects.
- Technical Insights: Gained experience in combining computer vision, natural language processing, and full-stack web development in a real-world setting.