Speech Engines and supported features
To learn what features and languages are supported by the different engines, use the following call:
GET /service/transcriber/engine
This API call lists all available languages for transcription with the specified engine.
Details
Description
Parameters
✘
Return Type
array[SpeechConfiguration]
Field Name | Required | Type | Description | Format |
---|---|---|---|---|
alignmentLanguageCodes |
✘ |
List of map |
The speech engine supports the transcript alignment feature on these languages. A list of objects with a code and label field. The code is the language code, the label is a more user-friendly label for the language. |
|
alignmentLanguages |
✘ |
List of string |
The speech engine supports the transcript alignment feature on these languages. A list of user-friendly labels for the language. |
|
billingType |
✘ |
String |
The speech engine will consume resources of this type. |
Enum: DATA, PROXY, CLOUD, PRODUCTION, NONE, ARCHIVED, DELETED, TRANSCRIPTION, KALDI, TRANSCRIPTION_3, TRANSCRIPTION_4, TRANSCRIPTION_5, TRANSLATION, FILERESOURCE, |
dictionarySupported |
✘ |
Boolean |
The speech engine supports the dictionary feature. |
|
engine |
✘ |
String |
The api reference of this speech engine. |
Enum: VOLDEMORT, KALDI, VOLDEMORT2, VOLDEMORT3, VOLDEMORT4, VOLDEMORT5, |
transcriptLanguageCodes |
✘ |
List of map |
The speech engine supports transcription on these languages. A list of objects with a code and label field. The code is the language code, the label is a more user-friendly label for the language. |
|
transcriptLanguages |
✘ |
List of string |
The speech engine supports transcription on these languages. A list of user-friendly labels for the language. |
Content Type
-
application/json
Responses
Code | Description | Datatype |
---|---|---|
0 |
default response |
List[ |
The response is a list of SpeechConfiguration objects, each describing a speech engine.
For example:
[
{
"engine": "VOLDEMORT2",
"billingType": "TRANSCRIPTION",
"dictionarySupported": false,
"transcriptLanguageCodes": [
{ "code": "ar", "label": "Arabic" },
{ "code": "nl", "label": "Dutch"},
{ "code": "en", "label": "English"},
...
],
"alignmentLanguageCodes": [
{ "code": "nl", "label": "Dutch" },
{ "code": "en", "label": "English" },
...
],
"transcriptLanguages": [ ... ],
"alignmentLanguages": [ ... ],
},
...
]
-
transcriptLanguageCodes
tells you which languages are supported for transcription by this engine -
If you want to use the alignment feature, look at
alignmentLanguageCodes
instead. Some engines don’t support alignment on all languages, or even not on any language! -
If you want to use the custom dictionaries feature, verify that
dictionarySupported
istrue
-
The
billingType
gives you information on which kind of resource your subscription needs to have. At the time of writing, only Enterprise plans support speech engines other than VOLDEMORT3.