Voice Control ESP32 with Sentiment Analysis

This project allows you to control an ESP32 device using voice commands and sentiment analysis. It consists of two Python scripts (speech_recog_using_API.py and speech_reg.py) and an Arduino sketch for the ESP32.

Prerequisites

Python Dependencies

• websocket
• speech_recognition
• requests

You can install these dependencies using pip: pip install websocket speech_recognition requests

Hardware Requirements

• ESP32 board
• Wi-Fi network credentials

Setup

1. ESP32 Setup:

   • Upload the provided Arduino sketch to your ESP32 board.
   • Update the ssid and password variables with your Wi-Fi network credentials.

2. Python Script Setup:

   • For speech_recog_using_API.py:
     • Update the server_address variable with the IP address of your ESP32 device.
     • Provide the URL of the sentiment analysis API in the sentiment_analysis_url variable.
   • For speech_reg.py:
     • Update the websocket_url variable with the WebSocket URL of your ESP32 device (e.g., ws://172.20.10.6:81).

Usage

1. With Sentiment Analysis (speech_recog_using_API.py):

   • Run the speech_recog_using_API.py script.
   • Speak into the microphone when prompted.
   • The script will perform sentiment analysis on the recognized text and send the corresponding "on" or "off" command to the ESP32 device based on the sentiment score.

2. Without Sentiment Analysis (speech_reg.py):

   • Run the speech_reg.py script.
   • Speak into the microphone when prompted.
   • If the script recognizes the word "on", it will send the "on" command to the ESP32 device, turning on the built-in LED.
   • If the script recognizes the word "off", it will send the "off" command to the ESP32 device, turning off the built-in LED.

Note

• Make sure that the ESP32 device and the computer running the Python scripts are connected to the same Wi-Fi network.
• For speech_recog_using_API.py, you need to provide a valid sentiment analysis API URL. If you don't have one, you can modify the script to use a different approach or remove the sentiment analysis part.