Youtube_Audio_To_Text

A simple python script to download audio and its subtitles from Youtube, then use google speech to text api to recognizi this audio.

Environment

python 3.6
os: Linux

Steps

There are only little steps to implement the function we want.

1. Run the following line in the command line

   export GOOGLE_APPLICATION_CREDENTIALS=[path]

   The path is the local file address you put service-account-file.json .

2. Change the uri in the main.py

   uri = 'https://www.youtube.com/watch?v=LspIeUElIFA'

   Here, you can put any Youtube video uri.

3. After running main.py, you will get a new dir, which contains audio.flac, google_result_audio.txt, subtitles.vtt. Then, you can compare the difference between the video subtitles and the text recognized by google speech to text api.

Example results

You can have a quick look at the audio.flac, google_result_audio.txt, subtitles.vtt here.

Suggested video uri

https://www.youtube.com/watch?v=LspIeUElIFA
https://www.youtube.com/watch?v=bR1oIe6m-Ds

The duration time of video should not be too long, because it will take more time for google api to recognize the audio.