search

RamtinMahdavifar / Resident-Care

0 stars 0 forks source link

readme

Resident-Care

Project Files

Access the project files and documentation through this Google Drive link.

Hardware Requirements

  1. 25Gb Free Disk Space
  2. 16Gb Ram
  3. Network Card or Ethernet port
  4. Stereo Output
  5. Microphone Input
  6. x86 CPU Architecture

Software Requirements

  1. Python 3.9.x (We have tested on Python 3.9.18)
  2. Debian Linux (We have tested on Ubuntu 22.0.4 Operating System)
  3. A Paid Tier 1 Open AI api key is required to run this Project.

Setup Instructions

Building and Installing Requirements

  1. Ensure that you have a version of python 3.9 installed on your system. \ We have used Python 3.9.18.

  2. Run the provided bash script setup_project.sh to install the required packages, and setup the virtual environment. 

    ./setup_project.sh

Vosk Setup (Local Voice Recognition)

  1. Download and extract a vosk model from Vosk Models. \ It is recommended to use vosk-model-en-us-0.42-gigaspeech.zip for the highest accuracy.
  2. Move the extracted folder to the root folder of the Python project.

Environment Configuration

Create a .env file in the project root with the following content:

# Open API
OPENAI_API_KEY='Your Open API Key'

# Twilio
ACCOUNT_SID='Your ACCOUNT_SID'
AUTH_TOKEN='Your AUTH_TOKEN'
TWILIO_PHONE_NUMBER='Twilio's phone number assigned to your account'
CAREGIVER_PHONE_NUMBER='The phone number of the caregiver who will receive alerts'

Ensure that the phone numbers are in full international format, starting with 
the '+' sign followed by the country code and the full phone number. This number must
be verified in your Twilio account to recieve messages.

For example:

In Canada, the country code is +1, so the format will be +1XXXXXXXXXX.
If you are using a phone number from another country, replace +1 with the 
appropriate country code for that number.

# Vosk
VOSK_MODEL_PATH='Path to your downloaded Vosk model'

# TTS
TTS_MODEL_NAME='tts_models/en/ljspeech/tacotron2-DDC_ph' 

You can replace the TTS Model with another if you prefer.

# Resident Details
These environment variables are used to personalize the experience based on the resident's details.

RESIDENT_FIRST_NAME='The first name of the resident'
RESIDENT_LAST_NAME='The last name of the resident'
RESIDENT_AGE_YEARS='The age of the resident in years (1-130)'
RESIDENT_SEX='The sex of the resident ("Male" or "Female")'
RESIDENT_MEDICAL_CONDITIONS='A comma seperated list of known medical conditions of the resident'

# Setting Resident Details
Set these variables to reflect the resident's personal information. 
Ensure that the age is an integer between 1 and 130, and the sex is 
specified as either "Male" or "Female" (case insensitive). 
The medical conditions should be a comma-separated list of 
conditions without any special characters.

For example:

RESIDENT_FIRST_NAME='John'
RESIDENT_LAST_NAME='Doe'
RESIDENT_AGE_YEARS=85
RESIDENT_SEX='male'
RESIDENT_MEDICAL_CONDITIONS='mild dementia, mobility issues.'

# CareGiver Details
Ensure that this is not empty.

CAREGIVERS_DESCRIPTION='The role of the Caregiver'

For example:

CAREGIVERS_DESCRIPTION='Is a nurse in charge of taking care of the patient.'

Running the Program

  1. Ensure you have completed the setup instructions above.
  2. Ensure your virtual environment is activated. If it is not activated run the following command. 
    source venv/bin/activate

Running as a Terminal Application

To run the program as a terminal application, use the following command: 

python3 /path/to/your/main.py

Running as a Web Application with a User Interface

To run the program as a web application with a user interface, use the command:

streamlit run /path/to/your/main.py

Please adjust the /path/to/your/main.py with the actual path to your main.py file on your system.