🖼️ Vision AI

A Flask-based API that generates captions for images using a custom deep learning model (BLIP). The API provides endpoints for:

• Generating image captions
• Generating captions with bounding boxes

🚀 Features

✅ Generate descriptive captions for uploaded or base64-encoded images ✅ Optionally overlay a bounding box on the image ✅ Fast processing time ✅ CORS-enabled for cross-origin access

---

📂 Project Structure

blip/
├── app.py                # Flask server
├── model.py              # Image captioning model logic
├── index.html            # Simple frontend for testing

---

🔑 API Endpoints

Endpoint	Method	Description
/	GET	API status and list of endpoints
/caption	POST	Generate a caption for an image
/caption_with_box	POST	Generate a caption and return image with a bounding box

POST request

• Form field: image (file upload or base64 string)

---

🛠 Installation

1️⃣ Clone the repository

git clone https://github.com/sebastianbenjamin/blip.git
cd blip

2️⃣ Install the required packages

pip install -r requirements.txt

---

🚀 Run the app

python app.py

By default, it will run at:

http://0.0.0.0:5000/

---

🌐 Frontend

You can test the API using the provided frontend:

Open blip/index.html in your browser

This allows you to upload images and see the generated captions directly.

---

📝 Example cURL requests

Generate caption

curl -X POST -F "image=@/path/to/your/image.jpg" http://localhost:5000/caption

Generate caption with bounding box

curl -X POST -F "image=@/path/to/your/image.jpg" http://localhost:5000/caption_with_box

---

📌 Notes

• Input can be an uploaded image file or a base64-encoded string in the image field.
• Response includes processing time in seconds.

---

📄 License

This project is licensed under the MIT License.

---

👉 Quick Start:

python blip/app.py

Then open:

blip/index.html

in your browser!

---