MindWave Project Setup Guide

The MindWave project focuses on detecting symptoms of mental health patients through the analysis of voice data. To get started with this project, follow these step-by-step instructions for setting up and running it in Visual Studio Code (VSCode).

Note: Ensure that the path to the project folder contains no spaces, as spaces in the path can lead to errors. Also, run VSCode as an Administrator for uninterrupted task execution.

1. Running Firmware Part

1.1 Install Communication Drivers

• Download the communication drivers for ESP32:

  • Driver Link: CP210x Universal Windows Driver
  • Extract and save the drivers to a folder.

• Connect the ESP32 device to your PC and open the Device Manager.

• Select the relevant COM port for the ESP32 and click "Update driver."

• Choose "Browse my computer for drivers" and provide the path to the folder where you extracted the drivers.

• The driver installation will be completed.

1.2 Install PlatformIO Extension in VSCode

• Install the PlatformIO extension in Visual Studio Code.

1.3 Upload Firmware to ESP32

• Open VSCode.
• Navigate to the firmware directory.
• Upload the firmware to the ESP32 device via USB.
• Wait until the upload finishes.
• Press the reset button on the ESP32.

1.4 Connect ESP32 to Wi-Fi

• On your PC, search for a Wi-Fi Access Point named "MindWave_testing."
• Click on it and choose the Wi-Fi router you want the ESP32 to connect to.
• Enter the router's password.
• The ESP32 will now connect to the selected Wi-Fi router.

1.5 Verify Connectivity

• Make sure both your PC and the ESP32 are connected to the same Wi-Fi router.

2. Python App Part

use python 3.9.0

use latest pip version (or upgrade pip)

install Microsoft visual studio with C++ development. Ensure you install this as it is a dependency of a package called 'fairseq'.

2.1 Set Up Python Environment

• Navigate to the mlapi directory.

• Create a virtual environment:

  • python -m venv venv

• Activate the virtual environment:

  • venv/Scripts/activate

• If activated successfully, you will see (venv) alongside the path in the terminal.

2.2 Install Required Dependencies

• Install the project's Python dependencies by running the following commands:

  • pip install -r requirements.txt
  • pip install googletrans==4.0.0-rc1
  • pip install https://github.com/marianne-m/brouhaha-vad/archive/main.zip

• Clone the SpeechBrain repository and prepare SpeechBrain:

  • git clone https://github.com/speechbrain/speechbrain.git
  • cd speechbrain
  • pip install -r requirements.txt
  • pip install --editable .

• Download the required Llama2 model from: Llama-2-7B-Chat-GGML Model

• Store the downloaded Llama2 model into a folder named "llama2" inside the mlapi directory.

• Download scream detection model from following URL if not found in "scream_detection_models" folder: https://drive.google.com/file/d/1CbtLptAF_RxJrbfdZczAwoAnCItDNFxR/view?usp=sharing

• place the model in a folder called "scream_detection_models" inside the mlapi directory

2.3 Permissions Setup

• Before running the Python app, ensure that your PC user account has write permissions to the "C:" and "D:" drives.

3. Running the Python App in Production (if streaming device available)

• Run the alice_stream.py for running server for alice

  • python alice_stream.py

• in a different terminal, Run the main.py script to run recording and analyzing:

  • python main.py

4. Running the Python App in test mode (using recorded files)

• Run the test.py with any audio file with .wav format as input

Now, you have successfully set up and run the MindWave project for detecting symptoms of mental health patients through voice data analysis.

Note: Always run VSCode as an Administrator to avoid any interruptions during the execution of tasks.

Enjoy using MindWave to make a positive impact on mental health!