| 摘要 |
The invention relates to a device for speech recognition, in particular for a navigation system (1), comprising a recognition unit (2, 3) for generating at least one word hypothesis from a speech sequence, in addition to a comparison unit (4), which creates one or several resultant words from the word hypothesis or hypotheses, using possible words that have been entered in a data memory. (5). The aim of the invention is to improve the efficiency of the speech recognition process. To achieve this, the recognition unit (2, 3) is designed to derive at least two key words from one or several sequences of speech and to generate several word hypotheses for the two or more key words and the comparison unit (4) creates resultant words in relation to the key words from the several word hypotheses by comparing the latter with the words entered in the data memory (5), taking into account the fact that the resultant words are related by context. The invention also relates to a method for speech recognition.
|
