The improvements are not impressive to be widely adopted. It's sort of known
that phase information has litte effect on human speech recognition rate. This
fact was explored everywhere from speech recognition to VoIP codecs.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Why is phase spectrum not used in feature extraction process ?
It's frequently used, there are quite many publications about phase, for
example:
http://www-i6.informatik.rwth-aachen.de/publications/download/445/Schlueter-
ICASSP-2001.pdf
The improvements are not impressive to be widely adopted. It's sort of known
that phase information has litte effect on human speech recognition rate. This
fact was explored everywhere from speech recognition to VoIP codecs.