Using Microsoft Agent's TTS in C #

Introduction to Microsoft Agent

The Microsoft Agent API can provide a display of cartoon characters, and it can also support speech recognition, so applications can respond to voice commands, and cartoon characters can respond to commands through synthesized speech, recorded audio signals, or text.

Requirements for using the Microsoft Agent

To use this technique, we must have the following components:

• Microsoft Agent Core Components

• Cartoon characters in Microsoft Agent ━━genie, Merlin, Robby and Peedy

• Microsoft speech API 4.0a run Time Library

• Microsoft speech recognition engine

· Lernout and Hauspie text-speech engine

All of the above components can be downloaded from the http://microsoft.com/products/msagent/downloads.htm.

Introduction to Speech technology

Text-Speech conversion refers to the computer to convert text information into synthetic speech for output, speech recognition refers to the computer can recognize the speaker said, accept the speaker's command and input data.

Both speech recognition and text-speech conversion require an associated engine, and almost all speech recognition engines convert incoming speech data into phonemes related to a particular engine, and these phonemes are converted to text that the application can use.

Text-speech conversion in two ways:

1, synthetic text-speech conversion

2, continuous text-speech conversion

Synthetic text-Speech conversion:

In the synthesized method of pronunciation, the engine processes each word and produces the phonetic phoneme of the word, and then these phonemes are transferred into a complex algorithm that mimics the human voice and generates speech.

Continuous text-Speech conversion Mode:

In continuous text-speech conversion, the engine processes text information and finds sentences, words, and phrases from a pre-recorded voice library, in which the generated speech is continuous.

programming interfaces for voice applications

The Microsoft Voice Application programming interface uses the OLE Component Object mode (COM) architecture under WIN32 (Windows 95, Windows NT), and Microsoft's agent architecture uses the Microsoft Voice Application programming Interface (SAPI) in the synthesized speech output. Also uses SAPI to support speech input (speech recognition SR or text-to-speech TTS). The Microsoft Agent defines the interfaces that allow applications to access their services, allowing applications to control role animations, support user input events, and specify the output mode.

Role window

In the Microsoft Agent application, the cartoon characters are displayed in their respective windows, which always appear at the top of the z-order. The user can drag the role's window through the left mouse button, and the image of the role moves with the pointer.

Talking Balloon Ring

In addition to the voice output, the animated character also supports the type of speech in the form of a cartoon, the characters speak, the words appear in the balloon, when the end of the balloon circle is gone.