Speech Recognition in Python using Google Speech API

Difficulty Level : Hard
Last Updated : 22 Oct, 2017

Speech Recognition is an important feature in several applications used such as home automation, artificial intelligence, etc. This article aims to provide an introduction on how to make use of the SpeechRecognition library of Python. This is useful as it can be used on microcontrollers such as Raspberri Pis with the help of an external microphone.

Required Installations

The following must be installed:

Python Speech Recognition module:
```
 sudo pip install SpeechRecognition 
```
PyAudio: Use the following command for linux users
```
sudo apt-get install python-pyaudio python3-pyaudio
```
If the versions in the repositories are too old, install pyaudio using the following command
```
sudo apt-get install portaudio19-dev python-all-dev python3-all-dev && 
sudo pip install pyaudio
```
Use pip3 instead of pip for python3.
Windows users can install pyaudio by executing the following command in a terminal
```
pip install pyaudio
```

Speech Input Using a Microphone and Translation of Speech to Text

Configure Microphone (For external microphones): It is advisable to specify the microphone during the program to avoid any glitches.
Type lsusb in the terminal. A list of connected devices will show up. The microphone name would look like this
```
USB Device 0x46d:0x825: Audio (hw:1, 0)
```
Make a note of this as it will be used in the program.
Set Chunk Size: This basically involved specifying how many bytes of data we want to read at once. Typically, this value is specified in powers of 2 such as 1024 or 2048
Set Sampling Rate: Sampling rate defines how often values are recorded for processing
Set Device ID to the selected microphone: In this step, we specify the device ID of the microphone that we wish to use in order to avoid ambiguity in case there are multiple microphones. This also helps debug, in the sense that, while running the program, we will know whether the specified microphone is being recognized. During the program, we specify a parameter device_id. The program will say that device_id could not be found if the microphone is not recognized.
Allow Adjusting for Ambient Noise: Since the surrounding noise varies, we must allow the program a second or too to adjust the energy threshold of recording so it is adjusted according to the external noise level.

Speech to text translation: This is done with the help of Google Speech Recognition. This requires an active internet connection to work. However, there are certain offline Recognition systems such as PocketSphinx, but have a very rigorous installation process that requires several dependencies. Google Speech Recognition is one of the easiest to use.

The Above steps have been implemented below:

#Python 2.x program for Speech Recognition
  
import speech_recognition as sr
  
#enter the name of usb microphone that you found
#using lsusb
#the following name is only used as an example
mic_name = "USB Device 0x46d:0x825: Audio (hw:1, 0)"
#Sample rate is how often values are recorded
sample_rate = 48000
#Chunk is like a buffer. It stores 2048 samples (bytes of data)
#here. 
#it is advisable to use powers of 2 such as 1024 or 2048
chunk_size = 2048
#Initialize the recognizer
r = sr.Recognizer()
  
#generate a list of all audio cards/microphones
mic_list = sr.Microphone.list_microphone_names()
  
#the following loop aims to set the device ID of the mic that
#we specifically want to use to avoid ambiguity.
for i, microphone_name in enumerate(mic_list):
    if microphone_name == mic_name:
        device_id = i
  
#use the microphone as source for input. Here, we also specify 
#which device ID to specifically look for incase the microphone 
#is not working, an error will pop up saying "device_id undefined"
with sr.Microphone(device_index = device_id, sample_rate = sample_rate, 
                        chunk_size = chunk_size) as source:
    #wait for a second to let the recognizer adjust the 
    #energy threshold based on the surrounding noise level
    r.adjust_for_ambient_noise(source)
    print "Say Something"
    #listens for the user's input
    audio = r.listen(source)
          
    try:
        text = r.recognize_google(audio)
        print "you said: " + text
      
    #error occurs when google could not understand what was said
      
    except sr.UnknownValueError:
        print("Google Speech Recognition could not understand audio")
      
    except sr.RequestError as e:
        print("Could not request results from Google 
                                 Speech Recognition service; {0}".format(e))

Transcribe an Audio file to text

If we have an audio file that we want to translate to text, we simply have to replace the source with the audio file instead of a microphone.
Place the audio file and the program in the same folder for convenience. This works for WAV, AIFF, of FLAC files.
An implementation has been shown below

#Python 2.x program to transcribe an Audio file
import speech_recognition as sr
  
AUDIO_FILE = ("example.wav")
  
# use the audio file as the audio source
  
r = sr.Recognizer()
  
with sr.AudioFile(AUDIO_FILE) as source:
    #reads the audio file. Here we use record instead of
    #listen
    audio = r.record(source)  
  
try:
    print("The audio file contains: " + r.recognize_google(audio))
  
except sr.UnknownValueError:
    print("Google Speech Recognition could not understand audio")
  
except sr.RequestError as e:
    print("Could not request results from Google Speech 
              Recognition service; {0}".format(e))

Troubleshooting

The following problems are commonly encountered

Muted Microphone: This leads to input not being received. To check for this, you can use alsamixer.
It can be installed using
```
sudo apt-get install libasound2 alsa-utils alsa-oss
```
Type amixer. The output will look somewhat like this
```
Simple mixer control 'Master', 0
  Capabilities: pvolume pswitch pswitch-joined
  Playback channels: Front Left - Front Right
  Limits: Playback 0 - 65536
  Mono:
  Front Left: Playback 41855 [64%] [on]
  Front Right: Playback 65536 [100%] [on]
Simple mixer control 'Capture', 0
  Capabilities: cvolume cswitch cswitch-joined
  Capture channels: Front Left - Front Right
  Limits: Capture 0 - 65536
  Front Left: Capture 0 [0%] [off] #switched off
  Front Right: Capture 0 [0%] [off]
```
As you can see, the capture device is currently switched off. To switch it on, type alsamixer
As you can see in the first picture, it is displaying our playback devices. Press F4 to toggle to Capture devices.
In the second picture, the highlighted portion shows that the capture device is muted. To unmute it, press space bar
As you can see in the last picture, the highlighted part confirms that the capture device is not muted.
Current microphone not selected as capture device:
In this case, the microphone can be set by typing alsamixer and selecting sound cards. Here, you can select default microphone device.
As shown in the picture, the highlighted portion is where you have to select sound card.
The second picture shows the screen selection for sound card
No Internet Connection: The speech to text conversion requires an active internet connection.

This article is contributed by Deepak Srivatsav. If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please write comments if you find anything incorrect, or you want to share more information about the topic discussed above.

Attention geek! Strengthen your foundations with the Python Programming Foundation Course and learn the basics.

To begin with, your interview preparations Enhance your Data Structures concepts with the Python DS Course. And to begin with your Machine Learning Journey, join the Machine Learning – Basic Level Course

My Personal Notes

Related Articles

Speech Recognition in Python using Google Speech API