method for computer speech recognition, speech recognition and control system for controlling a technical system and telecommunications equipment

摘要

The method involves determining which speech recognition rate is required for a speech-recognition application, and determining what information content of the characteristic vector components and how many components are the minimum requirement in order to achieve the speech recognition rate. The speech recognition is carried out using characteristic vectors with the number of components required to provide the determined information content. Independent claims are included for a control device and for a telecommunication device with the control device.