Hotel AI Feedback System

A lightweight web application that uses computer vision to track and analyze customer experiences. By monitoring an exit camera, the system detects customer emotions (Positive/Negative) in real-time, respects privacy by optionally blurring faces, and visualizes the feedback analytics in a beautifully designed web dashboard.

Features

• Live Analytics Dashboard: Real-time insights into customer sentiment, emotion distribution, and recent feedback logs.
• AI Emotion Tracking: Utilizes OpenCV and YOLO for real-time face detection and tracking.
• Privacy-First Design: Includes an option to blur faces in the live video feed to comply with privacy standards.
• Dual Storage Options: Toggle between local JSON-based storage (works offline) and Firebase for cloud-synced data.
• Camera Management: Scan and select different camera sources directly from the settings panel.

Technology Stack

• Backend: Python, Flask, Waitress
• Computer Vision: OpenCV, YOLO (Ultralytics / OpenVINO)
• Frontend: HTML5, Vanilla CSS (Glassmorphism design), JavaScript (Chart.js)
• Storage: Local JSON / Firebase Admin SDK

Setup & Installation

1. Install Dependencies

Make sure you have Python installed. Then, install the required packages:

pip install -r requirements.txt

2. Run the Application

Start the Waitress production server:

python app.py

The application will start on port 5000. You can access the dashboard by opening your browser and navigating to http://localhost:5000.

Project Structure

feedback/
├── app.py                 # Main Flask application entry point
├── requirements.txt       # Python dependencies
├── config.json            # Configuration file (Storage mode, camera source, privacy)
├── yolov8n.pt             # YOLOv8 nano model weights
├── core/                  # Core application logic
│   ├── feedback_engine.py # Maps emotion data to positive/negative feedback
│   ├── storage.py         # Abstraction layer for Local JSON & Firebase storage
│   ├── tracker.py         # AI face/object tracking logic
│   └── video_stream.py    # Camera management and OpenCV stream processing
├── data/                  # Local storage directory (if using local mode)
├── templates/             # HTML templates for the dashboard and settings
└── static/                # Static assets (CSS, JS, Images)

Configuration

Settings can be dynamically updated via the web interface (http://localhost:5000/settings), or by modifying the config.json file directly:

• storage_mode: "local" or "firebase"
• camera: Dictionary containing the camera "source" index (e.g., 0 for default webcam).
• privacy: Dictionary containing the "blur_faces" boolean toggle.