V-Blaze and V-Cloud Online Help

Removing Sensitive Information from Transcripts and Audio

V‑Cloud’s redaction options can remove potentially sensitive numeric information from both transcripts and audio:

  • When transcript purification is activated (by setting the scrubtext  option to true), all instances of sensitive numeric digits in output transcripts will be replaced by the hash sign (#).

  • When audio purification is activated (by setting the scrubaudio  option to true), all audio segments containing sensitive numbers are replaced by silence. When using audio purification without other customization options, results are returned as a ZIP archive containing transcripts and purified MP3 files.

You can use the scrubconf option to specify the audio output format that you want to receive and the default whitelists used when scrubbing.

For example, to submit the file sample1.wav for transcription with both transcript and audio purification activated, use a command like the following:

curl -F token=123e4567e89b12d3a456426655440000 \
     -F scrubtext=true \
     -F scrubaudio=true \
     -F file=@sample1.wav \
     https://host-name:17171/transcribe

The response to this POST will be a requestid that enables you to retrieve a results file which contains both sample1.json and a purified version of sample1.wav in MP3 format.