Create text transcription for a video or audio file. This outputs 3 files, 1 words.json file that has all of the words said along with their timing and confidence, 1 transcript.txt with just all the words read, and 1 captions.srt file which is the captions for that video.
This is a fork of adhikary97/Sharetape-Speech-To-Text.
$ pip3 install moviepy vosk
Choose language models from vosk model page and unzip them in ~/vosk/ or any directory to your liking. You can use the -m option to specify the exact path of the language model at runtime.
Video must be .mp4 or .mov
$ python main.py videoname.mp4
Video must be .wav
$ python main.py audioname.wav