23.1.2 program voskmulti.pl
In order to process a certain section in a database with speech recognition, there is the utility program voskmulti.pl dbname start end audio cpus. With this program the range start to end in a database can be used for speech recognition. If start and end values are omitted, the entire database is used. If a '0' is specified for audio, all video files are processed, if a '1' is specified, all audio files are processed, and if a '2' is specified, all audio and video files are processed. The cpus option can be used to specify how many processor cores (CPUs) are to be used for speech recognition. If the cpus parameter is not specified, either 4 CPUs or the number of cores specified with the installation will be used. With the script voskcheck.pl (same parameters) a recognition can be started on all not yet processed files.
The language of the recognition depends on the OCR definition or the first language specified in it, see 16. Currently possible are German, English, French, Italian, Spanish, Portuguese and Dutch. It is important to note that the desired language must be listed first in the OCR definition.