SAPI Technical Article

Home Page
Research-Lab's approach

Preface

This article is being written in response to the assignment to carry out the complete research on various issues related to SAPI  for Research-Lab's upcoming  product's campaign.

This article is going to cover the important SAPI  issues, services, products, service providers and Research-Lab's approach to SAPI.

Research is focused on the following points

SAPI...general definition

SAPI stands for Service Access Point Identifier. An address used at layer 2 to manage different data types for the same individual device connecting to the ISDN network. The SAPI and TEI together form the layer 2 address. 

In detail SAPI  is a programming standard that provides tools and components to speech recognition and text-to-speech software applications. Speech recognition applications allow a computer to respond to spoken commands or to write text in response to spoken words. Text-to-speech applications allow a computer to synthesize a voice that speaks words that are in a string or text file.
 


Architectural details

SAPI is a specification for both Engine and Application developers. Speech engine developers can make their engines SAPI compliant by implementing the Engine level interfaces. Application developers use the Application level interfaces provided by SAPI to access the features of the underlying engine. 


Advantages of SAPI

SAPI has following advantages 

  1. Standard Specification - Adoption enables compliance with several speech engines, without additional development effort.
  2. COM based API - Supports application development and scripting. No learning curve in terms of implementation.
  3. Oriented towards feature-rich desktop applications.

 


Disadvantages of SAPI
  1. Limited to Microsoft Platform.
  2. No direct server-side implementation - Default implementation limited to a single user deployment scenario. The interfaces are not designed for large portal development with multiple access points.
  3. Engine specific profiles - User Profiles can't be ported across speech engines, as SAPI does not define standards for profile storage. Thus, Training required for each user on each engine.

 


SAPI & applications

SAPI is applicable for the interfaces of Speech Recognition and Text To Speech applications

Speech Recognition
Each speech recognition application requires speech recognizer object for encapsulating recognition engine.

Text To Speech
SAPI exposes a single interface for its TTS functionality, ISpVoice. The engine vendor supplies its TTS engine with a set of one or more voices. These voices can be characterized using a set of attributes such a Male/Female, Adult/Child, and language supported by the voice. When created, the ISpVoice object is ready to use with the default voice specified under the Control Panel's Speech settings dialog. Once created, however, the voices can be changed. SAPI provides methods to query for available voices on the engine.


SAPI Implementation Consideration

You can include the Speech API feature in your platform by choosing the feature from the Catalog. Choosing the Speech API feature sets the appropriate environmental variable for enabling the feature.

The following table shows the Sysgen variables that enable the Speech API feature.

Sysgen variable

Description

SYSGEN_SPEECH When this variable is set, the Speech API feature is included in the OS image.

When this variable is not set, the Speech API feature is excluded from the image.

 

SYSGEN_SPSRENG When this variable is set, the speech recognition engine feature is enabled.

Research topics are as follows

sapi 
sapi 5.1 
microsoft sapi 
sapi 5 
sapi 4.0 
sapi potong 
php sapi 
sapi 5.0 
sapi sdk 
sapi download 
sapi dll 
karapan sapi 
daging sapi 
tapi sapi 
speech sapi 
ternak sapi 
ms sapi 
microsoft sapi 5.1 
peternakan sapi 
sapi cli php 
sapi microsoft 
sapi voice 
susu sapi 
mapi sapi tapi 
sapi 5.1 download 
penggemukan sapi 
delphi sapi 
sapi engine 
sapi server 
sapi tts 
sapi h 
sapi php 
sapi voices 
sapi 4.0 a 
sapi 4 x 
sapi delphi 
microsoft sapi 4.0 
sapi tutorial 
sapi vb 
sapi text to speech 
mapi sapi and tapi 
sapi venezuela 
tapi mapi sapi 
jandi sapi 
mapi tapi sapi 
sapi cgi php 
microsoft sapi 5 
sapi spvoice 
windows sapi 
sapi 5.1 sdk 
sapi php4apache dll 
kulit sapi 
sapi bali 
sapi plates 
download sapi 5.1 
sapi s7 
gambar sapi 
sapi 4 download 
visual basic sapi 
sapi 5 download 
microsoft sapi download 
sapi gila 
what is sapi 
peternakan sapi perah 
microsoft sapi 4 
sapi treiber 
isdn sapi 
mapi sapi and tapi developer's guide 
sapi cpl 
sapi lib 
sapi srl 
sapi cli php error 1 
sapi server receiver 
sapi speech engine 
sapi 5 voices 
microsoft sapi 5.0 
peternakan sapi potong 
sapi 4.0 download 
sapi gmbh 
sapi island 
sapi net 
tapi sapi mapi 
sapi utan 
sapi 日本語 
download the microsoft sapi 4.0 a runtime binaries 
microsoft speech sapi 5.1 
sapi paper 
sapi selco 
sapi talker 
microsoft sapi 5.1 download 
sapi plate 
sapi tutorials 
sapi sdk download 
command failed for target sapi cli php 
sapi 4 sdk 
sapi spa 
inca sapi 
download sapi 5 
sapi server reciever 
sapi と は 
sapi ca 
sapi 4.0 a runtime binaries 
sdk5 1 
sdk 
speech 
5.1 
tts 
truvoice 
lernout & hauspie 
sapi5 
sapi4 
tts3000 
l&h tts3000 
spchapi 
speechapi 
lernout hauspie 
speech recognition 
jsapi 
digalo 
l&h 
spvoice 
lhttsspe 
msagent 
lernout 
lernout&hauspie 
synthesis 
vorlesen 
lhttsjpj 
sprachausgabe 
ttsapp 
screen reader 
viavoice 
text 
sapi4sdk 
speechsdk 
asr 
dragon naturally speaking 
logox 
api4 
sdk5 
speechsdk5 
narrator 
api 
jaws 
spracherkennung 
msapi 
cloudgarden 
dragon naturallyspeaking 
microsoft 
voice 
sprachsteuerung 
spracheingabe 
screenreader 
naturally speaking 
webspeech 
nitroderm 
dragon naturally 
computerstimme 
window eyes 
jandi 
screen readers 
sprachprogramm 
speechstudio 
viseme 
via voice 
could not be initialized 
langid 
agtx0c0a 
engine 
freedom scientific 
application programming interface 
xlisten 
vorlese 
object library 
hauspie 
command & control 
henter joyce 
voices 
sprachsoftware 
sprachprogramme 
software development kit 
sprachsynthese 
msttsa22l 
dragon dictate 
spracherkennungssoftware 
ispvoice 
tomreader 
dr abuse 
jfw 
jtts 
4.0 
command and control 
studienauswahl 
sprechen 
control panel 
diktiersoftware 
vtext 
truetalk 
wav2txt 
in3 
naturallyspeaking 
agent 
rext 
speak