Revision history for Perl extension Speech::Recognizer::SPX

0.09 Mon Jan 7 15:20:52 EST 2008

• Switch to using PocketSphinx rather than Sphinx2.
• Update everything for that.
• Backward compatibility with Sphinx2 has been mostly removed.

0.08 Fri Aug 5 16:35:37 EDT 2005

• Official release with N-best, segmentation.

0.07_92 Mon Jul 18 10:30:18 EDT 2005

• Incorporate proper backwards-compatibility code.

0.07_91 Tue Jul 12 12:31:51 EDT 2005

• Fix compilation/test on platforms without audio
• Try to be backwards-compatible with Sphinx2 0.3 (untested)
• Add support for word segmentation in uttproc
• Add support for N-best lists, funded by SingleTouch Interactive, Inc (http://www.singletouch.net/)

0.07 Fri Jun 3 12:23:46 EDT 2005

• Long-overdue update
• Restores compatibility with Sphinx2 0.5
• Makes it possible to use a non-standard --sphinx-prefix
• New contact addresses, etc.

0.06_02 Tue Mar 27 12:17:45 EST 2001

• First externally released version.
• No code changes from 0.06_01, just different copyright notices, with the addition of a README.

0.06_01 Fri Feb 9 12:52:49 EST 2001 (UNRELEASED)

• Give a way to pass raw data into Audio::SPX::Continuous (hooray for undocumented features :P)

0.06 Thu Jan 25 14:15:00 EST 2001 (UNRELEASED)

• Skipped a version, oh well.
• Add the sub-directories' test suites to MANIFEST (doh)
• Fix the calling conventions for uttproc_rawdata() and uttproc_result().
• Add a testsuite for Speech::Recognizer::SPX itself.
• Add demos of using Audio::SPX::Continuous with select().
• Add proper path searching for Sphinx2 libraries in Makefile.PL and use the exported $SPHINXDIR variable for defaults.
• Fix 'undefined value' warnings in Audio::SPX read functions.
• Allow the use of hash references for fbs_init().

0.04_99 Wed Dec 13 17:02:59 EST 2000 (UNRELEASED)

• Split off feature extraction and audio modules, link with necessary libraries
• Changed name to Speech::Recognizer::SPX to comply with the Sphinx2 license and fit into the CPAN namespace better
• Add exporter functionality to Speech::Recognizer::SPX, and put together some nice tags of things to export.
• The only thing missing before this can be released is documentation and a test suite, I think.

0.04 Mon Dec 4 17:34:34 EST 2000 (UNRELEASED)

• Add cepstral analysis (Sphinx2::FE) methods.
• Make Sphinx2::Audio::init and Sphinx2::Audio::Continuous::init class methods.
• Fix tty-continuous.

0.03 Sun Dec 03 14:17:46 2000 (UNRELEASED)

• Add a server module and a demo of streaming audio to Sphinx over the network. Remove the really broken and stupid pipe-demo.pl :)

0.02_01 Fri Dec 01 11:29:46 2000 (UNRELEASED)

• Add copyright notices.

0.02 Fri Dec 01 00:42:57 2000 (UNRELEASED)

• Fix the interface, removing evil autoload.
• Add the option to specify one's own ARGV to fbs_init()
• Add a hacked-up continuous-audio-from-random-file API.
• Add two demos.

0.01 Wed Nov 29 12:25:19 2000 (UNRELEASED)

• original version; created by h2xs 1.19