Fall Detection Model

Introduction

This project focuses on detecting falls using sensor data. The model processes data from various sources, including BLE sensors, and uses machine learning techniques to identify fall events accurately.

Table of Contents

• Introduction
• Project Structure
• Setup Instructions
• How It Works
• References

Project Structure

• dags/: Directed Acyclic Graphs for orchestrating tasks.
• hyperparameters/: Hyperparameter configurations for model training.
• src/: Source code for data processing and model training.
• tests/: Unit tests for the project.
• .coveragerc: Configuration file for measuring code coverage.
• .gitignore: Specifies intentionally untracked files to ignore.
• .isort.cfg: Configuration for sorting imports.
• .pre-commit-config.yaml: Configuration for pre-commit hooks.
• AirflowREADME.md: Instructions specific to Airflow setup.
• Dockerfile: Docker configuration file.
• Makefile: Commands for setting up and managing the project environment.
• README.md: Project overview and instructions.
• cdk.json, cdk_json.txt: AWS CDK configuration files.
• docker-compose.yml: Docker Compose configuration.
• pass-role-policy.json: AWS policy for passing roles.
• poetry.lock: Lock file for Poetry dependencies.
• postgres.env: Environment variables for PostgreSQL.
• pyproject.toml: Configuration file for project dependencies.
• requirements.txt: List of project dependencies.
• start_airflow.sh: Script to start Airflow.
• trust-policy.json: AWS trust policy configuration.

Setup Instructions

Prerequisites

• Python 3.10 or later
• AWS CLI: Ensure that your AWS CLI is configured with the necessary credentials.
• Node.js and npm: Required for AWS CDK.
• Docker: For containerized deployment.

Steps

1. Clone the repository:

   git clone https://github.com/eliaselhaddad/Fall-Detection-Model-Internship2.git
   cd Fall-Detection-Model-Internship2

2. Set up the virtual environment (optional but recommended):

   python -m venv env
   source env/bin/activate  # On Windows use `env\Scripts\activate`

3. Install dependencies using Poetry:

   pip install poetry
   poetry install

4. Set up AWS CDK:

   npm install -g aws-cdk
   cdk bootstrap

5. Set up Docker:

   docker-compose up

How It Works

The Fall Detection Model processes sensor data from BLE sensors, performs data preprocessing, and trains a machine learning model to detect falls. The workflow involves the following steps:

• Data Retrieval: Collects data from BLE sensors.
• Data Processing: Cleans and processes the data for model training.
• Model Training: Trains a machine learning model using the processed data.
• Event Handling: Uses AWS services like Lambda and CDK for managing events and infrastructure.

Example Commands

• Run the application:

  python -m src.app

• Process and predict sample data:

  python -m src.processing.source_all_processor --use_sample

References

• AWS CDK
• Poetry
• Boto3
• Pandas