AI Image Captioning Web App

This is a Streamlit-based web application that uses a BLIP (Bootstrapped Language-Image Pre-training) model for generating captions based on user-uploaded images. The app allows users to upload images in JPG, JPEG, or PNG formats, and it provides an AI-generated caption for the uploaded image.

Features

• Image Upload: Users can upload images directly into the app.
• Image Display: The uploaded image is displayed on the screen.
• AI Captioning: The app uses the BLIP model to generate and display a caption for the uploaded image.
• Developer Credits: Displays developer profiles with customizable information in the footer.

Requirements

The following Python libraries are required to run the application:

streamlit==1.26.0 Pillow==10.0.0 transformers==4.33.0 torch==2.0.1

Install all dependencies using the following command:

pip install streamlit Pillow transformers torch

How to Run the Application

1. Clone the repository or download the code.

2. Install the required packages by running the command:

   pip install -r requirements.txt

3. Run the Streamlit app by executing the following command in the terminal:

   streamlit run app.py

4. Open your browser and go to the address:

http://localhost:8501

or the URL provided by Streamlit after running the app.

Code Structure

• app.py: Contains the main logic of the application.
• requirements.txt: Lists all required Python libraries.
• README.md: Documentation for the project (this file).

Model

• The app uses the BLIP Image Captioning Model from Hugging Face's model hub (Salesforce/blip-image-captioning-base) to generate captions based on uploaded images.

Developer Credits

Instagram : KARTHIK PRABHU LinkedIn : KARTHIK PRABHU karthikprabhu.netlify.app

Customization

• You can easily customize the developer details in the footer section by updating the profile card HTML and the LinkedIn URLs in the app.py file.

Future Improvements

• Add more styling to enhance user experience.
• Include support for multiple languages in caption generation.
• Optimize image processing speed for larger images.