Offline Voice Control on Raspberry Pi.

Designed To Run On:

• Raspberry Pi.
• Raspberry Pi OS.
• Python 3.

The Voice Control Consists of:

• Speech-to-Text engine is a piece of software that interprets human voice into a string of text. It lets the computer know what is being said.
• Text-to-Speech engine converts text into sound. It allows the computer to speak, probably as a response to your command.

The Voice Control uses:

• Vosk as the Speech-to-Text engine.
• PYTTSX3 as the Text-to-Speech engine.

Why Vosk ?

• Supports 20+ languages and dialects - English, Indian English, German, French, Spanish, Portuguese, Chinese, Russian, Turkish, Vietnamese, Italian, Dutch, Catalan, Arabic, Greek, Farsi, Filipino, Ukrainian, Kazakh, Swedish, Japanese, Esperanto, Hindi, Czech, Polish. More to come.
• Works offline, even on lightweight devices - Raspberry Pi, Android, iOS.
• Installs with simple pip3 install vosk.
• Portable per-language models are only 50Mb each, but there are much bigger server models available.
• Provides streaming API for the best user experience (unlike popular speech-recognition python packages).
• There are bindings for different programming languages, too - java/csharp/javascript etc.
• Allows quick reconfiguration of vocabulary for best accuracy.
• Supports speaker identification beside simple speech recognition.

to use the Voice Control System you will need to:

• Speakers Plugged into the Raspberry Pi usingn Jack.
• Microphone Plugged ( Wired USB or Bluetooth ).

Let's Get Started.

• First you need to install the necessary library packages which are:

  • pip install pyaudio.
  • pip install vosk.
  • pip install pyttsx3.

• or by using pip install -r requirments.txt

• After checking that the Speaker and Microphone works properly start using the Voice Control System using the Engine implementation.

Speech To Text Engine:

Start using by importing Voice() which has seconds parameters that you can specity, default is 10 in your main application.

Code Implementation:

def voice(seconds = 10):
   start_time = time.time()
  
   stream.start_stream()
   print("Model Started ....")
   
   while True:

       current_time = time.time()
       elapsed_time = current_time - start_time
       
       data = stream.read(4000,exception_on_overflow = False)
       recogniser.AcceptWaveform(data)

       result = recogniser.Result()[14:-3]
       print(result)

       if elapsed_time > seconds:
           break

voice()

Text To Speech Engine:

Start using by importing speak() in your main application.

Code Implementation:

def speak(audio):
   engine = pyttsx3.init()
   engine.setProperty('rate', 150)
   print("Assistant : " + audio)
   engine.say(audio)
   engine.runAndWait()

Note: If you use the code in another python script you need to import these libraries.

import pyttsx3
from vosk import Model, KaldiRecognizer
import pyaudio
import os
import time

It's that simple!