CODEWITHSUNDEEP
pythonperformancespeech-recognitionpocketsphinx
Colby Grieves
Colby Grieves

Reputation: 53

Python Speech Recognition - Sphinx

I am make a simple speech recognition program that will enable me to control my robot with voice command. I only want the program to look for certain words and be relatively fast. My project is based on Micheal Reeves 'I made a robot that shines a laser in my eye' and am trying to create something similar to the voice commands seen in his video.

The issue I am having is that sphinx is fast, but (EDIT: NOT ACCURATE). As well as this, when I enable keywords, the out goes weird. If I say command shutdown, the output will be :

"three  nine  one  four  five  eight  two  one  eight  nine  three  four  two  six  zero  eight  nine  two  one  six  four  eight  seven  one  three  four  nine  five  eight  two  eight 
four  five  nine  three  one  two  eight  six  nine  three  five  seven  two  zero  one  nine  five  eight  two  four  four  nine  one  five  eight  three  two  six  four  two  zero  seven  one  nine  three  four  five  eight  two  five  one  three  four  eight  two  six  eight  zero  one  three  four  five  two  seven  eight  eight  three  nine  five  two  four  eight  one  two  eight  two  eight  two  eight  command shutdown  command  eight  one  four  three  eight  two  two  eight "

I am not sure to fix this and I tried to do recognise_google but and it was much more accurate but really slow and i want the keywords enabled so that it only checks if a said a collection of words and then prints it the the screen if i did.

The other issue I am having is with the listen_in_background() function. I cant seem to get it working properly.

Here is my code:

import speech_recognition as sr
import pocketsphinx

keywords = [
    ("command", 1), 
    ("one", 0), 
    ("two", 0), 
    ("three", 0), 
    ("four", 0), 
    ("five", 0), 
    ("six", 0), 
    ("seven", 0), 
    ("eight", 0), 
    ("nine", 0), 
    ("zero", 0), 
    ("command x axis add", 0), 
    ("command y axis add", 0), 
    ("command x axis subtract", 0), 
    ("command y axis subtract", 0), 
    ("command clear shift string", 0), 
    ("command shutdown", 0),
    ("command flip tracking", 0), 
    ("command pause", 0), 
    ("command detect face", 0), 
    ("command detect body", 0)
]

def speech2text():
    r = sr.Recognizer()
    with sr.Microphone() as source:
        r.adjust_for_ambient_noise(source)
        audio = r.listen(source) #this is were i want to listen in the background to run it at the same 
        #time as other code
    try:
        data = r.recognize_sphinx(audio, keyword_entries = keywords)
        return data
    except:
        return "Error..."

while True:
    print(speech2text())

Upvotes: 3

Views: 3504

Answers (1)

Zac
Zac

Reputation: 1

I had the same issue. I tried different sensitivities from 0 to 1 and found that if all keywords have a sensitivity of over 0.9, they recognize equally and reasonably accurately and don't randomly spam the output phrase. If the value is any lower than that it spits way too many keywords than is reasonable.

I also got an UnknownValueError when ANY word that wasn't a keyword. If you're looking for a way to only detect those keywords, I'd definitely try setting their sensitivities all to 1 and seeing where that gets you. I think the only downside might be that if words in the keyword list are similar than you might get different hits than you'd expect.

Upvotes: 0

Related Questions