AI-Driven Cultural Artifact Recognition App

Overview

This project is an AI-powered web application designed to recognize and classify cultural artifacts from uploaded images. The system uses deep learning algorithms to match artifacts from a dataset and returns relevant details, such as the artifact’s name, history, and significance, sourced from a CSV file (artifact.csv). The project is built using Flask as the web framework and serves as a practical application of Data Augmentation in Deep Learning for Computer Vision.

Features

• Image Upload & Recognition: Users can upload images of artifacts, and the system will match and classify them based on a pre-trained model.
• Artifact Details: Once an artifact is recognized, relevant information is fetched from a CSV file and displayed.
• Augmented Dataset: The dataset of artifacts includes augmented images for improved classification accuracy.
• Multilingual Support: The system provides QR codes and labels in multiple languages for ease of access.
• Webcam Capture: Users can capture images via webcam directly on the website for recognition.
• API Endpoint: An API endpoint is provided for external systems to classify artifacts by sending images.

Architecture

• Flask: The web framework used for serving the application.
• Machine Learning: A deep learning model (likely CNN or ResNet) is used for artifact recognition.
• Dataset: The artifact.csv file contains metadata about the artifacts (name, description, etc.).
• HTML/CSS/JavaScript: For building the frontend with AR and webcam features.
• Nginx & Gunicorn/Waitress: For serving the Flask application in production.

Tech Stack

• Backend: Python, Flask, Gunicorn (or Waitress for Windows)
• Frontend: HTML, CSS, JavaScript
• Machine Learning: TensorFlow/Keras or PyTorch (based on your model)
• Database: CSV for artifact data storage
• Deployment: Nginx, Gunicorn (Linux), Waitress (Windows)

Prerequisites

• Python 3.8+
• Flask
• Gunicorn (for Linux) or Waitress (for Windows)
• TensorFlow/Keras or PyTorch
• Nginx (for Linux server setup)

Setup Instructions

1. Clone the Repository

git clone https://github.com/yourusername/ArtifactRecognitionApp.git
cd ArtifactRecognitionApp

2. Install Dependencies

pip install -r requirements.txt

3. Setup the Dataset

You can download the dataset from the following Google Drive link:

Download Dataset from Google Drive

Once downloaded, place the dataset (CSV and images) in the data/ folder. Ensure that the artifact.csv file is properly formatted with the correct metadata (e.g., artifact name, description, etc.).

4. Running the Application Locally

To run the application locally using Flask’s built-in development server:

python app.py

Alternatively, if using Waitress:

waitress-serve --port=8080 app:app

5. Deploying to a Production Server

On Linux (with Gunicorn and Nginx):

• Install Nginx:

  sudo apt-get install nginx

• Start the Flask app using Gunicorn:

  gunicorn --bind 0.0.0.0:8000 app:app

• Configure Nginx as a reverse proxy for Gunicorn (see nginx.conf for configuration).

On Windows (with Waitress):

• Start the Flask app using Waitress:

  waitress-serve --port=8080 app:app

6. Procfile for Deployment (Heroku)

Create a Procfile for Heroku deployment:

web: waitress-serve --port=$PORT app:app

Example Use Cases

1. Cultural Artifact Classification: Upload a picture of an ancient artifact (e.g., statue, pottery) and receive detailed information about its historical significance.
2. AR Support: Scan QR codes at museums or historical sites to get additional information on artifacts in multiple languages.
3. Webcam Integration: Capture images of artifacts using a webcam for instant recognition.

Folder Structure

.
├── app.py                  # Main Flask application
├── artifact.csv            # CSV file with artifact details
├── data/                   # Dataset directory
├── templates/              # HTML templates
├── static/                 # Static files (CSS, JS)
├── requirements.txt        # Python dependencies
├── Procfile                # Heroku Procfile for deployment
└── README.md               # Project documentation

Future Enhancements

• Implement additional language support.
• Add support for larger datasets by integrating a more robust database (e.g., PostgreSQL).
• Improve the machine learning model with more complex architectures.

License

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