bensilence 0.1.4

A Python library for recording voice using voice actvity detection (VAD). Best for AI voice assistants.

This is my first repsitory. So, I may not have done everything properly. I expect your understanding.

bensilence

A voice recorder with voice activity detection (VAD). It starts recording when VAD detects speech and stops recording as the speaker stops talking for a second. Best use case would be AI assistants. This is just another version of rhasspy-silence. This library uses Cobra (Picovoice) for VAD, which is better than webRTC that was used in rhasspy-silence.

Installation

1. Install the package:

   pip install bensilence 
   

Install via GitHub

1. Clone the repository:

   git clone https://github.com/benimrans/bensilence.git
   cd bensilence
   

2. Install dependencies:

   pip install -r requirements.txt 
   

3. Setup:

   py setup.py install 
   

Usage/Examples

You'll need a Picovoice account to get an API key. Once you get it, you are ready to go!

from bensilence import silence

silence = silence(api_key="your_picovoice_api_key")

silence.initialize()

result, file_name = silence.record()

print(result, file_name)

Here are some variables that you may want to change:

These are the default values.

before_seconds = 1 # Adds the unrecorded parts to the recording by going back 1 second from the time the speech started.
silence_threshold = 1 # Stops recording if there's silence for 1 second. 
max_sensitivity = 2 # {1: 0.5, 2: 0.7, 3: 0.9} Value that VAD should detect to start recording.
max_recording_time = 30 # Maximum recording time.

Related

This is the reason why I made this library:

rhasspy-silence

License

MIT