How does one define a keyword spotter ? One way to look at it could be an
engine that processes one complete
utterance from start to end (like pocketsphinx for example) and checks if the
keyword was spoken anywhere in that
utterance.
The other way to look at it is as an engine which keeps listening to the sound
continuously (for however lengthy periods of times) and anytime a keyword is
spoken it identifies it... so basically it always listens and hunts for a
keyword.
Which of the above will be termed as keyword spotter in speech recognition
jargon and which of the schemes (or
some other) is supported by say Sphinx-4 ?
Thanks and regards,
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
Both of them are keyword spotters. A keyword spotter looks for the keyword in
an utterance, it is independent of when it reports it's finding.
Both styles of implementation are possible to implement in sphinx-4
(here is one such implementations).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
How does one define a keyword spotter ? One way to look at it could be an
engine that processes one complete
utterance from start to end (like pocketsphinx for example) and checks if the
keyword was spoken anywhere in that
utterance.
The other way to look at it is as an engine which keeps listening to the sound
continuously (for however lengthy periods of times) and anytime a keyword is
spoken it identifies it... so basically it always listens and hunts for a
keyword.
Which of the above will be termed as keyword spotter in speech recognition
jargon and which of the schemes (or
some other) is supported by say Sphinx-4 ?
Thanks and regards,
Hi,
Both of them are keyword spotters. A keyword spotter looks for the keyword in
an utterance, it is independent of when it reports it's finding.
Both styles of implementation are possible to implement in sphinx-4
(here is one such implementations).