Skip to content

creates text from audio of A/V input file, using docker, sphinx. extracts keywords and NLP entities. leverages OpenNews, Stanford, Oxford, CMU and more

Notifications You must be signed in to change notification settings

traceypooh/audio2text

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

audio2text

docker-based pocketsphinx that takes A/V input file to create text from the audio and extract keywords and entities from the results.

Allows one to, once SETUP is done, run locally on a Mac, network disabled / no cloud, and run speech to text and text analysis on an A/V file.

PREREQUISTES

Docker (eg: https://docs.docker.com/docker-for-mac/install/ )

git (brew or XCode setups have you covered ;-)

SETUP

git clone https://github.com/traceypooh/audio2text.git
cd audio2text
docker build -t audio2text .

RUN

( docker run --rm -i audio2text |tar xf - ) < test.mp3

will make (click each to see the results):

  • out.json - detailed word/phrase with timings
  • out.txt - transcript of entire audio/video file
  • out.srt - timed transcript of audio/video file
  • out.key - keywords extracted from .txt (above)
  • out.plo - Persons, Locations, Organizations (and more) extracted from .txt (above)

About

creates text from audio of A/V input file, using docker, sphinx. extracts keywords and NLP entities. leverages OpenNews, Stanford, Oxford, CMU and more

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages