Changes

Revision history for SailAlign

1.20   Friday, August 30th 2013
        + Third release, removed dependency from the voice activity detection module, which has been
        found to not affect performance. The audio file is now segmented in chunks of equal length. 
        Possible breaks within words are expected to be taken care of in the succeeding iterations. 
        + Released on github now.
        + Removed the acoustic models from the release file. These are now downloaded from a website.
        + Updated the README to also include information on how to cope with the possible Audio::Wav
        warning about ambiguous call to function. A patch is provided to correct Audio::Wav::Read.pm

1.10    Wednesday, June 29th 2011
        + Second release, thanks to Jangwon Kim for pointing out some problematic
        cases of forced word alignment.
        + Out-of-vocabulary words are treated as silence for forced word alignment
        (which can optionally run after the iterative algorithm). 
        + Minor bugfixes.

1.00    Saturday, January 8th 2010    
        + First version, released on an unsuspecting world.