X-Ray Feature Backend

Table of Contents

• Introduction
• Features
• Prerequisites
• Installation
• Directory Structure
• Configuration
• Usage
• API Endpoints
• Data Flow Overview
• Troubleshooting
• Additional Notes
• License

Introduction

This backend application replicates a simplified version of Amazon Prime Video's X-Ray feature. It processes uploaded videos to recognize faces in different scenes and identifies known actors using facial recognition and classification models. The application provides a FastAPI-based API for video uploading, processing, and retrieving processed results.

Features

• Face Detection: Detects faces in video scenes using face_recognition and OpenCV.
• Face Recognition: Matches detected faces against a database of known actors.
• Scene Processing: Processes videos based on provided scene timestamps.
• API Endpoints: FastAPI application with endpoints for video uploading and results retrieval.
• Data Storage: Stores actor information and processing results in JSON files.
• Extensibility: Modular code structure for easy expansion and customization.

Prerequisites

• Python: Version 3.7 or higher.
• CMake: Required for building dlib, which is a dependency of face_recognition.
• Git: For cloning the repository (optional).

Installation

1. Clone the Repository

   git clone https://github.com/your-username/your-repo-name.git
   cd your-repo-name

2. Create a Virtual Environment

   python -m venv venv
   source venv/bin/activate  # On Unix or macOS
   venv\Scripts\activate  # On Windows

3. Install Dependencies

   pip install --upgrade pip
   pip install -r requirements.txt

4. Install CMake (if not already installed)

   • Windows: Download from the official CMake website
   • Ubuntu/Debian: sudo apt-get install cmake
   • macOS: brew install cmake

Directory Structure

project/
├── main.py
├── models/
│   ├── face_recognition.py
│   └── face_classification.py
├── data/
│   ├── known_actors/
│   ├── actors.json
│   ├── known_faces.pkl
│   ├── uploaded_videos/
│   └── results/
├── requirements.txt
├── .gitignore
└── README.md

Configuration

1. Prepare Known Actors Data
2. Create actors.json
3. Generate Known Faces Database

For detailed configuration steps, refer to the full documentation.

Usage

1. Set up the Known Actors Database
2. Run the FastAPI Server:

   uvicorn main:app --reload

3. Test the API endpoints

For detailed usage instructions, refer to the full documentation.

API Endpoints

• POST /upload_video
• GET /results

For detailed API documentation, refer to the full documentation.

Data Flow Overview

1. Video Uploading
2. Face Detection and Recognition
3. Results Generation
4. Response to Client
5. Cleanup

Troubleshooting

• CMake Not Found Error
• dlib Installation Issues
• Face Recognition Accuracy
• Empty Results
• Server Errors

For detailed troubleshooting steps, refer to the full documentation.

Additional Notes

• Security Considerations
• Performance Optimization
• Extending the Application
• Data Privacy

License

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

Acknowledgments

• Libraries Used: FastAPI, Uvicorn, face_recognition, OpenCV, Pydantic
• Inspiration: Amazon Prime Video's X-Ray feature

Feel free to contribute to this project by submitting issues or pull requests. If you have any questions or need further assistance, please contact the project maintainer.