simil.art is an application for Content-Based Image Retrieval (CBIR) based on deep features extracted from neural networks, specifically ResNet18 and fine-tuned ResNet50 on an artistic dataset. The purpose of this project is to work with an artistic dataset, retrieve similar images, and refine the results based on both the style of the query image (using fine-tuned models) and the color (using color histogram correlation).

• Content-Based Image Retrieval (CBIR)
• Vizualise feature maps
• Vizualise t-SNE plot
• Refinement of results based on color histogram correlation
• Refinement of results based on the style

To get started with the simil.art project, follow these steps:

1. Clone the repository:

   git clone https://github.com/yassinefkh/simil.art.git

2. Install the required dependencies:

   pip install -r requirements.txt

3. Download the large files containing feature vectors and Faiss indexes using Git LFS:

   git lfs pull

To run the application, execute the following commands:

 python manage.py

Then, open a web browser and navigate to http://localhost:8000.

Upload an image to search for similar images.

This project uses data from WikiArt for educational purposes only and is not intended for commercial use.

• Project Author: FEKIH HASSEN Yassine, CALMANOVIC-PLESCOFF Auguste, HIBAOUI Imane, MARTIN--SAVORNIN Alain
• Project Supervisor: Pr. KURTZ Camille