speech2text

Speech to Text command line utility based on SpeechRecognition Python library

install dependencies

pip3 install -r requirements.txt

How to run it

python3 speech2text.py [OPTIONS] input-file.wav [output-file.txt]

options

--silence-length length, -l length : Split sentences when silence is length ms. or more. Default: 700 ms

--silence-thresh amount, -t amount : Consider silent if quitter than amount dBFS. Default: -40

--temp-dir path, -T path : Path to store temporary files. Default: ./tmp

--padding padding -p padding : How much silence to add at both ends so audio slice doesn’t seem abruptly cut. Default: 100 ms.

--language language, -L language : Language to use for speech recognition. Default: en-US

--ambient-noise, -n : Try to mitigate ambient noise. Not always improves conversion. So off by default

--keep-temporary, -k : Do not remove temporary files created during conversion. Useful for debugging

--silent, -s : Do not show progress on console

--max-workers workers, -m workers : Number of concurrent conversions threads. Default: number of OS reported CPUs.