Speech recognition is performed only for new incoming audio and video files. To automate
existing documents, the internal script vosk2db.pl dbname doc
in the directory /home/cvs/archivista/jobs
can be used.
The program vosk2db.pl
always works with one processor unit (CPU). To run the speech recognition subsequently over a whole database or a range of it, the program voskmulti.pl
can be used.