| 摘要 |
Voice control of a service application provided to a terminal from a remote server is distributed between the terminal and a remote application part. A relatively low power automatic speech recognition system (ASR) is provided in the terminal for recognizing those portions of user-supplied audio input that relate to terminal functions or functions defined by a predefined markup language. Recognized words may be used to control the terminal functions, or may alternatively be converted to text and forwarded to the remote server. Unrecognized portions of the audio input may be encoded and forwarded to the remote application part which includes a more powerful ASR. The remote application part may use its ASR to recognize words defined by the application. Recognized words may be converted to text and supplied as input to the remote server. In the reverse direction, text received by the remote application part from the remote server may be converted to an encoded audio output signal, and forwarded to the terminal, which can then generate a signal to be supplied to a loudspeaker. In this way, a voice control mechanism is used in place of the remote server's visual display output and keyboard input.
|
