| 摘要 |
Apparatus for processing audio data to provide processed audio data to a speech search engine has a number of processing elements (61, 70, 72, 74, 80, 82, 90, 92, 10) connected by a number of buffers (60, 62, 71, 73, 75, 100, 81, 83, 91, 93). Some buffers (60, 93) receive data as processed, whereas others only store data when requested by the processing element connected to the buffer output. A user may design speech recognition apparatus by selecting processing elements and buffers from a library. |
