ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
0

pocketsphinx recognizer.py has initial incorrect output

asked 2013-12-20 09:06:35 -0500

dan gravatar image

updated 2014-01-28 17:18:53 -0500

ngrennan gravatar image

I am using pocketsphinx and it works great for recognition. However, when recognizer.py first starts, it outputs one or two random words from the dictionary, as if there was a queue that has to be cleared. This happens even if the microphone is muted. Any idea where that might be originating? Is there some way to suppress that?

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
1

answered 2013-12-21 22:44:31 -0500

fergs gravatar image

I've not seen this before -- but I imagine it is likely an issue with either gstreamer or the VADAR gstreamer plugin which is used to segment the audio input. There's currently no built in options to handle such a problem, but it could probably be easily added by storing the time at which the pipeline was enabled and making sure that some time as passed when an utterance arrives.

edit flag offensive delete link more

Comments

OK, I'll try that.

dan gravatar image dan  ( 2013-12-23 13:23:57 -0500 )edit

Question Tools

2 followers

Stats

Asked: 2013-12-20 09:06:35 -0500

Seen: 261 times

Last updated: Dec 21 '13