The speech mapping defines which messages from a SAPI and JSAPI speech recognition and synthesis engine correspond to a particular Hub message. Those messages that do not have a corresponding Hub message are not currently supported. If such support is required, please send mail to darpa-hub@linus.mitre.org.
The following speech mapping message exist:
Speech Recognition
Speech Synthesis
SAPI Speech Recognition
Engine
|
|
|
JSAPI Speech Recognition
Engine
|
|
|
||
Current Speaker | Not implemented | Send a message to request a notification of who the current speaker is in a speaker dependent recognizer. | ||
Delete Rule | Not implemented | Sends a message to remove a rule from the grammar. | ||
Dictation Context | Not implemented | Provides context information of what words can follow or precede a given word. Used for dictation grammars. | ||
Dictation Correction | Not implemented | Provides dictation correction information after recognition. | ||
List Dictation Words Added | Not implemented | Request that the recognizer send a message that contains a list of the new words added to the dictation grammar. | ||
List Dictation Words Removed | Not implemented | Request that the recognizer send a message that contains a list of the words removed from the dictation grammar. | ||
List Known Speakers | Not implemented | Requests that the recognizer send a message that contains a list of the known speakers in a speaker dependent recognizer. | ||
List Rules Names | Not implemented | Request that the recognizer send a message that contains a list of the rule names for the current grammar. | ||
Recognizer Information | Not implemented | Request that the recognizer send recognizer specific information. | ||
Recognizer State | Not implemented | Request that the recognizer send information about the recognition state. | ||
Result Token | Not implemented | Request that the recognizer send Result Token notifications, which indicate how to change the spoken text into written form (for dictation grammars). | ||
Rule | Not implemented | Request that the recognizer send information about a given rule name. | ||
Rule Active | Not implemented | Make a particular rule active in the recognition engine. | ||
Speaker Profile | Not implemented | Defines a speaker profile for a speaker dependent recognizer on the recognition server. | ||
Notification Messages |
||||
---|---|---|---|---|
Notification Current Speaker | Not implemented | A notification message from the recognition server that indicates the name of the current speaker. Primarily used for a speaker dependent recognizer. (The message is sent in response to the Current Speaker message.) | ||
Notification Dictation Words Added | Not implemented | A notification message from the recognition server that returns a list of words added to the dictation grammar. (The message is sent in response to the List Dictation Words Added message.) | ||
Notification Dictation Words Removed | Not implemented | A notification message from the recognition server that returns a list of words removed from the dictation grammar. (The message is sent in response to the List Dictation Words Removed message.) | ||
Notification Grammar Name | Not implemented | The recognition server issues the name of the grammar currently loaded. | ||
Notification Known Speakers | Not implemented | A notification message from the recognition server that returns a list of known speakers stored within the recognizer. (This message is typically used for a speaker dependent recognizer.) (The message is sent in response to the List Known Speakers message.) | ||
Notification Listener | Not implemented
(Need to verify this one!!) |
The recognition process sends this message to indicate whether a message has been recognized. Possible return values are: 'accepted result', 'recognition update', 'rejected result', released audio', 'released correction info'. | ||
Notification Recognizer Information | Not implemented | The recognition process this message to provide information about the speech recognition engine itself. This message is sent in response to the Recognizer Information message. | ||
Notification Recognizer State | Not implemented | The recognition issues this message to indicate the state of the recognition engine. The possible states for the recognizer are: 'recognizing', 'accepted', or 'rejected'. This message is sent in response to the Recognizer State message. | ||
Notification Result Token | Not implemented | The recognition process issues this message to indicate how to change the spoken text into written form (for dictation grammars). This message is sent in response to the Result Token message. | ||
Notification Rule | Not implemented | The recognition process return information about a grammar rule. The information contains the rule status, rule name, and rule. The message is sent in response to the Rule message. | ||
Notification Rule Names | Not implemented | The recognition process rules a list of rule names in the current grammar. The message is sent in response to the List Rule Names message. | ||
Recognition Results | Not implemented | Request that the recognition serer provide information about the recognition results. The type of information that can be requested are: n_best, threshold, sentence, samples, status, rule, speaker, vocabulary, and clear. | ||
Recognition Results Speaker | Not implemented | The recognition process provides information about the speaker such as the speaker mode, speaker name, gender, and age. This message is sent in response to the Recognition Results Request message. | ||
Recognition Results Vocabulary | Not implemented | The recognition process provides information about the grammar such as size, perplexity, grammar, and grammar type. This message is sent in response to the Recognition Results Request message. |
|
|
|
Control Tag Translation | Not implemented | Given a control tag, the synthesis will sound a particular emotion or voice. Only bookmark is defined. We have not defined messages for: emotion, language, and speaking style. |
Get Status | Not implemented | Requests that the synthesizer send status values for pitch, real time, speed, and volume. |
Pitch | Not implemented | Change the pitch, after reinitialize. |
Real Time | Not implemented | Change the real time setting, after reinitialize. |
Volume | Not implemented | Change the volume stetting, after reinitialize |
Notification Status | Not implemented | The status response from the synthesis in response to the Get Status message. Returns the current values for pitch, real time, speed, and volume. |
|
|
|
Cancel | Not implemented | Cancel the output at the top of the queue. |
Cancel All | Not implemented | Cancel all output in the synthesizer queue. |
Pitch | Not implemented | Change the pitch, after reinitialize. |
Pitch Range | Not implemented | Change the pitch range level. |
Speaking Rate | Not implemented | Change the average speaking rate in words per minute, after reinitialize. |
Status | Not implemented | Request a status notification. |
Volume | Not implemented | Change the volume, after reinitialize. |
Notification Synthesizer Status | Not implemented | Will not receive the status notification, which indicates the pitch, pitch range, speaking rate, volume, voice name, voice gender, and age. |