Speech ExpressTM Architecture
The Speech Express architecture is shown by the diagram below.
Host Processor: The host processor is a microprocessor or microcontroller that controls the user interface and the DSP.
SpOSTM Driver: This provides the interface between the host applications and the speech processing applications on the DSP.
Using the SpOS Driver API instead of "glue code" greatly simplifies the process of upgrading to a different host processor,
different DSP or different speech processing applications on the DSP.
SpOSTM: SpOS is a component of the Speech Express firmware suite. It is Amai Technologies' DSP operating system for
speech processing. Its scalability makes it configurable for simple systems with only one DSP application as well as for more
complicated systems with as many as 16 DSP applications. The command processor and the codec driver are the required components
of SpOS. Additional more advanced operating system features can be added to a particular system if needed. This scaleable
architecture ensures that valuable ROM is not wasted on modules that are not required.
DSP Speech Applications: The DSP speech applications are the other component of the Speech Express firmware suite.
Only the DSP applications desired by the client are included in the ROM ready module. For example, a system might have only a
single speech codec for record and playback and a text-to-speech engine. SpOS is configured to include only those resources
required by the included applications. Speech applications are not limited to those offered in the Speech Express firmware
suite. Amai Technologies will also port custom applications to SpOS.
Analog-Digital Codec: The Analog-Digital Codec converts digital signals from the DSP to analog signals to be played back
on headphones or speakers. It also converts the analog signals from the microphone to digital signals to be processed by the
DSP. The codec driver module of SpOS provides the interface between the codec and the DSP speech applications.