[
https://issues.apache.org/jira/browse/STANBOL-1007?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13925749#comment-13925749
]
Rupert Westenthaler commented on STANBOL-1007:
----------------------------------------------
This Issue is about implementing a Stanbol EnhancementEngine [1] that uses CMU
Sphinx to transfrom spoken language into plain text. This means an engine that
consumes an audio file parsed with the ContentItem [2]; extracts the text from
the audio file and adds the extracted text as "plain/text" Blob to the same
ContentItem.
The engine will also need to support loading of "Acoustic and Language Models"
via the Apache Stanbol DataFileProvider [3] infrastructure.
This issue needs to be seen in context to STANBOL-1291 that describes a
possible Phonetic Linking enhancement chain for Apache Stanbol.
[1] http://stanbol.staging.apache.org/docs/trunk/components/enhancer/engines/
[2] http://stanbol.staging.apache.org/docs/trunk/components/enhancer/contentitem
[3] http://stanbol.staging.apache.org/docs/trunk/utils/datafileprovider
> Speech to Text Enhancement Engine using CMU Sphinx
> --------------------------------------------------
>
> Key: STANBOL-1007
> URL: https://issues.apache.org/jira/browse/STANBOL-1007
> Project: Stanbol
> Issue Type: Sub-task
> Components: Enhancer
> Reporter: Rupert Westenthaler
> Labels: gsoc2014, mentoring
>
> CMUSphinx is a speaker-independent large vocabulary continuous speech
> recognizer released under BSD style license
> (http://cmusphinx.sourceforge.net/wiki/)
> This library could be used to add Speech to Text capabilities to Stanbol. It
> is important that Enhancement Results keep track of the temporal position of
> the extracted text within the processed media file.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
