Training process and use of a speech recognition system, speech recognizer and training system

摘要

A method for training and operation of a speech, in particular for speaker-independent speech recognition, comprising a plurality of speech samples data bases, training formed an overall model, wherein at least a portion of the speech sample data bases is formed in coordination with each having a predetermined typical operating environment of the speech and the speech is trained in a relatively independent training phases for obtaining references separately with these speech samples data bases.