CMU Sphinx
From Wikipedia, the free encyclopedia
CMU Sphinx sometimes simply known as Sphinx is the general term to describe a group of speech recognition projects. These includes a series of speech recognizers (Sphinx I - IV), acoustic model trainer (SphinxTrain).
In 2000, the Sphinx group at CMU committed to open source several speech recognizer components, including Sphinx II and later Sphinx III (in 2001). Therefore, CMU Sphinx could also mean a group of open source projects related to speech recognition. For most, the speech decoders are perhaps the most well-known.
CMU Sphinx is perhaps the only open source, large vocabulary, continuous speech recognition project which consistently releases its work under the liberal BSD-license.
As for the components of CMU Sphinx, currently it is available in four forms.
Contents |
[edit] Sphinx 2
The fastest recognizer among all in the project, its focus is on real-time analysis of speech signals. It is widely used in dialogue systems and language learning systems. It can be used in computer based PBX systems, such as Asterisk.
[edit] Sphinx 3
A slower, but more accurate version of sphinx, used for batch processing. Recently it could also be used for recognition with gramars -N gram.
[edit] Sphinx 4
A speech recognition engine written in Java.
[edit] PocketSphinx
An embedded version of Sphinx which could be used in ARM.
[edit] External links
 |
This software-related article is a stub. You can help Wikipedia by expanding it. |
