ArtLens

ArtLens logo

ArtLens

ITE 351 · AI & Applications — Group project blog

Team · Proposal


Task 1 · Group members

Name Organization Email
Carayon Theophile Computer Sience theophile.carayon@edu.ece.fr
Picard Arthur Computer Sience arthur.picard@edu.ece.fr
Gozutok Leyan Computer Sience leyan.gozutok@edu.ece.fr
Malavaux Mikael Computer Sience mikael.malavaux@edu.ece.fr

Task 3 · Research idea

Title:

ArtLens - Exploring Artistic Movements Using AI and Machine Learning

Proposal:

Our research idea is to develop ArtLens, a web application that helps users learn more about paintings by simply uploading a photo. The goal is to make art history more accessible, especially for people who do not know the artist, title, or artistic movement of a painting.

The project will combine image recognition and machine learning techniques. We plan to use a pretrained CLIP model with PyTorch to compare uploaded images with a dataset of paintings collected from museum databases. A multimodal AI model, such as Gemini, will then analyze visual characteristics like colors, brushstrokes, lighting, and composition to suggest possible artistic movements.

The application could also identify visually similar paintings and provide information about their artists, collections, and historical sale prices when reliable records are available.

We will initially focus on five artistic movements and evaluate the system using a small dataset of labeled paintings. The main objective is not perfect recognition accuracy, but to demonstrate how AI and machine learning can be integrated into a practical and educational web application.