The whole process of building speech recognition systems is highly computationally expensive and time consuming. Fortunately as a Python Programmer, we don't have to worry about any of this! A number of speech recognition services are available for use online through an API and many of these also offer Python SDK's.
-
The SpeechRecognition library stands out in terms of ease-of-use among many other available packages.
-
The SpeechRecognition library acts as a wrapper for several popular speech APIs and is thus extremely flexible. One of these—the Google Web Speech API—supports a default API key that is hard-coded into the SpeechRecognition library. That means we can get off your feet without having to sign up for a service.
Before running the recorder.py script, please make sure to install the following libraries:
- SpeechRecognition
pip install SpeechRecognition
- pyaudio
pip install pyaudio
The recorder.py file takes in two optional arguments:
- --show_all : Whether to show all the possible transcripts.
[Default: False] - --pre_process : Whether to pre-process the data to tacke noise.
[Default: False]
An example command to run the program:
python recorder.py -show_all True -pre_processs True