A speech-to-text system for quick, cost free transcription
  What to expect when requesting

How does it work?

  • 1
    Download and run the debian package
  • 2
    Upload video/speech files into the tool, using the instructions provided
  • 3
    Edit the transcription, as required

What is Kaldi?

Kaldi is a speech recognition toolkit, built upon the open source software originally developed for use by speech recognition researchers. The quickest way to search through a piece of audio or video is via a transcript, but transcription by hand is a costly and time consuming endeavour. Kaldi has been designed as a means of automating the same process for free, requiring only a small amount of installation effort from a software developer. The BBC-Kaldi component provides a machine learning model built using the tools in the Open Source Kaldi toolkit and audio / text data from BBC programme and an easy to use interface that makes it simple to get up and running.

Top tips to get you ready

What have we learned from Kaldi?

Kaldi has been used extensively within the BBC, with a variety of learnings from each project. BBC Newslabs' Window On the Newsroom project showed that speech-to-text is a tremendous timesaver for journalists who need to be across a wide number of video feeds. Their Audiogram project proved that creating subtitles for viral videoclips can be done much quicker than was previously imagined. BBC Rewind also used speech-to-text to open up almost a million hours of material from the BBC Archive.

Latest Discussions

No discussions on Kaldi yet. Why not start one?