SnapInsight

This project is a web application that allows users to upload images, generate captions using a pre-trained BLIP model, and display the images along with their captions in a gallery. The application is built using Gradio for the web interface and utilizes the transformers library for image captioning.

Features

• Upload images from your device or capture using a webcam (if on a laptop).
• Generate captions for uploaded images using the BLIP model.
• Display images and their captions in a gallery.
• Save images and captions to a JSON file for persistence.

Requirements

• Python 3.7+
• gradio
• Pillow
• torch
• transformers
• user_agents

Installation

1. Clone the repository:

   git clone https://github.com/bnarasimha/SnapInsight.git
   cd SnapInsight

2. Create a virtual environment and activate it:

   python -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install the required packages:

   pip install -r requirements.txt

Usage

Method 1

1. Run the application:

   python app.py

2. Open your web browser and navigate to http://0.0.0.0:7860.

3. Upload an image or capture one using your webcam (if on a laptop).

4. Click the "Submit" button to generate a caption and add the image to the gallery.

Method 2

You can also run the application by executing run.sh file under scripts folder by providing just the droplet name, huggingface token and ssh fingerprint

cd scripts  
chmod +x run.sh
./run.sh

File Structure

• app.py: Main application file.
• requirements.txt: List of required Python packages.
• saved_images/: Directory where uploaded images are saved.
• image_data.json: JSON file where image data (filenames and captions) are stored.

License

This project is licensed under the MIT License. See the LICENSE file for details.