As you might have noticed there are practically no speech tools yet in the UCMA 2.0 Speech SDK apart from some basic classes.
Currently Microsoft is prioritizing which tools to make available in the Wave ‘14’ release of OCS/UCMA, at which time the Speech Server functionality should become an integral part of the UC platform.
Microsoft would like to solicit the feedback from the GotSpeech.net and the GotUC.Net Communities on how to prioritize which tools to build and ship for Telephony Speech (for Desktop Speech, please check out the Windows Vista Dictation Resource Kit)
This is specifically about the speech recognition and speech synthesis tools and not about the dialog languages, like VoiceXML and Windows Workflow. UCMA 2.0 shipped with basic UC Workflow Activities including speech, and currently the Speech Server (2007) VoiceXML 2.1 browser is ported to the UCMA API and this browser will be released in the future.
Using the comment tool, could you please distribute $100 in value across the following tools we are considering to build/release, with $0 being not important at all, and any value above $0 giving it a relative value increasing with importance. You can just copy the table below into a comment or if you prefer you can just copy and email it to me – I’ll see that it gets to the right parties.
|
Tool
|
Relative value to you out of $100 total
|
Comments
|
|
Grammar Static Analyzer
|
|
Syntax checking
|
|
Grammar Static Analyzer
|
|
Grammar cop like in Speech Server (2007) (detecting OOV utterances)
|
|
Grammar Dynamic Analyzer
|
|
Simulator to measure grammar changes effect on the basis of existing log files
|
|
Custom Pronunciation Tool (text)
|
|
Create custom pronunciations
|
|
Custom Pronunciation Tool (audio)
|
|
Generate pronunciations from audio
|
|
Lexicon Editing Tool (app)
|
|
Tool to edit custom application lexicon
|
|
Lexicon Editing Tool (master)
|
|
Ability to edit the Microsoft provided master lexicon
|
|
Prompt Editor
|
|
Edit and align prompts in prompt database for prompt engine
|
|
Prompt Database
|
|
Store prompts, ability to check missing prompts and do version control
|
|
Prompt Engine
|
|
Concatenative prompt engine
|
|
Speech Synthesis Prompt Sculpting
|
|
Tool to edit TTS pronunciations
|
|
Custom TTS Voice Tool
|
|
Tool to create your custom TTS voice
|
|
Visual Grammar Editor
|
|
Visual SRGS grammar editor
|
|
Statistical Grammar compiler
|
|
Tool to import transcriptions and create a statistical language model from those
|
|
Conversational Grammar Engine
|
|
Conversational grammar engine like in Speech Server (2007)
|
|
Conversational Grammar Builder
|
|
Conversational grammar builder like Speech Server (2007)
|
|
Transcription Tool
|
|
Tool to transcribe and annotate the logs in a consistent manner
|
|
Log Analysis Tools
|
|
Analyze log files to see hit rates and success metrics for dialog states like in Speech Server (2007)
|
|
Offline Recognition Tool
|
|
Ability to read wave files and measure recognition rates
|
|
ASR Analysis tools
|
|
Measures word error rates, sentence error rates, etc.
|
|
Other – please specify
|
|
|
As per Microsoft’s default policy, this posting is provided "AS IS" with no warranties, and confers no rights.
Many thanks in advance for your feedback.
Posted
05-04-2009 8:40 AM
by
marshallharrison