Skip to content
Snippets Groups Projects
Select Git revision
  • master default
1 result

stt

  • Clone with SSH
  • Clone with HTTPS
    • Name Last commit Last update
    cmd
    .gitlab-ci.yml
    Dockerfile
    README.md
    dispatch.go
    go.mod
    go.sum
    http.go
    model.go
    recognition.go
    README.md

    Stream To Text

    This tool uses speech recognition to transcribe audio streams in real time. It uses VOSK for recognition, supporting many different languages (though you have to specify the language upfront).

    The tool operates as a HTTP server, with a single endpoint accepting url and lang query parameters, e.g.:

    https://localhost:3242/?url=https://s.streampunk.cc/ondarossa.ogg&lang=it

    the transcribed response will be returned as a sequence of HTTP/1.1 chunked response parts.