Home › Forums › OpenEars › Changing Sphinx parameters

This topic has 2 voices, contains 2 replies, and was last updated by  Joseph S. Wisniewski 242 days ago.

Viewing 3 posts - 1 through 3 (of 3 total)
Author Posts
Author Posts
September 16, 2011 at 3:25 pm #7608

Joseph S. Wisniewski

What is the least painful way to change sphinx parameters from an OpenEars using application? I want to be able to do things like setting -hmm, -rawlogdir, -fwdflat -bestpath, -tmat etc. right before starting a recognizer, not as compile-time options.

Right now, I’m torn between using argfiles (which is a bit tricky on an iOS system, since you need to write bundle paths for the -lm -fsg -mdef etc parameters, and adding a second commandArray that can be passed to listeningLoopWithLanguageModelAtPath
September 16, 2011 at 4:17 pm #7609

Halle

Hi Joseph,

I would definitely do this by having a second commandArray.
September 19, 2011 at 10:22 pm #7610

Joseph S. Wisniewski

Thanks.

I ended up implementing both. ;)

I’ve found it’s convenient to have a Sphinx arguments file accompanying each batch of raw speech files, so I can use that as a baseline in tuning.

Viewing 3 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic.