Sphinx is a speaker-independent large vocabulary continuous speech recognizer software. It is also a collection of open source tools and resources that allows researchers and developers to build speech recognition systems.
It is released under the BSD license.
First, you *must* have SphinxBase, which you can download from http://cmusphinx.org/. You should download and unpack it to the same parent directory as PocketSphinx, so that the configure script and project files can find it. On Windows, you will need to rename 'sphinxbase-X.Y' (where X.Y is the SphinxBase version number) to simply 'sphinxbase' for this to work.
In a unix-like environment (such as linux, solaris etc):
· build and optionally install SphinxBase. If you want to use fixed-point arithmetic, you *must* configure SphinxBase with the --enable-fixed option.
· if you downloaded directly from the CVS repository, you need to do this at least once to generate the "configure" file:
· if you downloaded the release version, or ran "autogen.sh" at least once, then compile and install:
% make clean all
% make test
% make install
What's New in This Release:
· New, re-entrant API
· GStreamer plugin
· Python module with access to most of the API
· (slow) support for continuous density models
· more flexible semi-continuous modeling
· Reduced code size
· Reduced memory footprint
· Up to 18% faster on ARM platforms
· Word posterior probabilities (confidence scores) are available
· Experimental JSGF grammar file support
· New default WSJ language model, more suitable for dictation