SmartMeet - Intelligent Meeting Support System

Project Description

SmartMeet is an intelligent system designed to enhance meeting experiences through speech detection, idea generation, and late arrival detection. With advanced features such as real-time transcription, automatic summary generation, and keyword extraction, SmartMeet transforms the way meetings are conducted.

Authors

• Lucía Cordero
• Jorge Garcelán
• Laura González

Course: Ambient Intelligence

Date: January 2025

Technologies Used

• Python 3.9+
• Flask (Backend and API)
• Transformers.js and Whisper (Speech recognition and real-time transcription)
• OpenAI API (Summary, idea, and keyword generation)
• YOLOv10 and COCO-SSD (People detection in meetings)
• Jinja2 and HTML/CSS/JS (User interface)

Key Features

🔹 Real-Time Transcription and Visualization

• Uses Whisper with Transformers.js for live meeting transcriptions.
• Updates transcriptions every 3 seconds to balance speed and accuracy.
• Optimized formatting with speaker identification.
• Supports Spanish, English, and Italian.

🔹 Summary, Keyword, and Idea Generation

• Automatic summaries generated with OpenAI API.
• Keyword extraction based on natural language processing.
• Idea generation based on detected discussion topics.
• Interactive idea visualization through movable cards organized by priority.

🔹 New Participant Detection

• YOLOv10 implementation for precise participant detection.
• COCO-SSD as an alternative for faster processing.
• Grace period system to alert about late arrivals.

🔹 Additional Features

• Adaptive interface: Light/dark mode and font size adjustments.
• Timeline: Chronological log of generated ideas.
• Final report: Automatic generation of a complete meeting summary.
• Language selection: Multilingual support for interface and transcription.

Project Setup

Creating a Virtual Environment

1. Navigate to your project directory:

   cd /Users/jorgegarcelan/Desktop/UNI/5-MASTER IAA/SC2/AMBIENTAL/IAA-Aml/PROJECT

2. Create a virtual environment:

   python3 -m venv .venv

3. Activate the virtual environment:

   • On Windows:

     .venv\Scripts\activate

   • On macOS and Linux:

     source .venv/bin/activate

4. Install the required packages:

   pip install -r requirements.txt

Project Components

Web Component

The web component is a Flask application that serves the frontend of the project. It is configured using the config.ini file.

• File: web/web.py
• Configuration:
  • WEB_HOST: Host for the web server
  • WEB_PORT: Port for the web server
  • DEBUG: Debug mode for the web server
  • SECRET_KEY: Secret key for the Flask application

To run the web server:

python web/web.py

API Component

The API component is another Flask application that serves as the backend API for the project. It is also configured using the config.ini file.

• File: api/api.py
• Configuration:
  • API_HOST: Host for the API server
  • API_PORT: Port for the API server
  • SECRET_KEY: Secret key for the Flask application

To run the API server:

python api/api.py

Database Configuration

The database configuration is specified in the config.ini file under the [DATABASE] section.

• Configuration:
  • HOST: Host for the database server
  • PORT: Port for the database server
  • USERNAME: Username for the database
  • PASSWORD: Password for the database
  • DATABASE_NAME: Name of the database

Ensure that your database server is running and accessible with the provided configuration.

Configuration File

The config.ini file contains all the necessary configurations for the web, API, and database components. Make sure to update this file with your specific settings.

[DEFAULT]
SECRET_KEY = 0000000011111111

[API]
HOST = localhost
PORT = 5000

[WEB]
HOST = localhost
PORT = 8080
DEBUG = True

[DATABASE]
HOST = localhost
PORT = 5432
USERNAME = your_username
PASSWORD = your_password
DATABASE_NAME = your_database_name