Offline speech recognition and transcription to SRT file
-
Download at least one vosk model from alphacephei
I suggest vosk-model-en-us-0.22 to maximise the result
Once downloaded a vosk model, unzip the file and rename the folder to "model", then place it in the project's root directory. -
You also may need ffmpeg, if it is not present in your system, place ffmpeg executable in the project's root folder.
-
Vosk has been already compiled as dynamic library in his official repository
download it and place libvosk.so under lib/ and you are ready to go!
> cd src
> make
./mp4toSRT sample.mp4
the result will be in result.srt