Where's My Pic?

Where's My Pic? is a local image search engine powered by CLIP (Contrastive Language-Image Pre-training). It allows you to search for images in your local directories using natural language queries.

You can give it a try on Lightning AI!

Features

• Seamless integration with CLIP for powerful image-text retrieval.
• Easily index and update your local image directories.
• Perform natural language searches to find relevant images.
• Gradio-based user interface for a smooth searching experience.

Some Sneak Peeks

Demo Video

Check out the demo video to see Where's My Pic? in action:

Getting Started

1. Clone the repository:

   git clone https://github.com/Om-Alve/Wheres_My_Pic.git
   cd Wheres_My_Pic

2. Install the required dependencies:

   pip install -r requirements.txt

3. Index your image folders:

   python build-db.py --add /path/to/image/folder1 /path/to/image/folder2

   This will create a Chroma database in the img_db directory for the first time it's executed and index the specified folders. You can use it to add new folders.

4. To add the changes to the indexed folders, for example adding or deleting a file use the following command

   python build-db.py --update

   This will update the existing database with the changes in the indexed folders.

5. Run the application:

   python app.py

   This will launch the Gradio interface, where you can enter natural language queries to search for images.

Usage

1. Enter your search query in the text box (e.g., "a meme about machine learning").
2. The application will display the top results matching your query.
3. You can click on the images to view them in more detail.

Contributing

Contributions are welcome! If you find any issues or have suggestions for improvements, please open an issue or submit a pull request.

License

This project is licensed under the Apache License.