SPEAKER RECOGNITION
The presentation aimed at giving the general view of how human beings are able to recognize someone’s voice by hearing in relatively to the machine how can identify two different voices. Speaker recognition is the process of automatically recognizing speaking basing on the recorded voice. The lesson was very pleasant as the clarification was quite easy to understand. The speaker recognition is of two kinds which are speaker verification and speaker identification
The system works under the principle of pitch variation, the sound pitch which is subjected into disparities is recorded baring into mind that the sound is in analogous signals. The interesting issue is how the analog speech is converted into digital signal then upon comparison of current speech and stores the previous speech.
The speaker recognition are used for sake of security purpose where the user has to do several trial of recording their voices so the machine is just taking the average of sound pitch and stores it. Once the user comes and wants to user the machine would repeat the same words which will be recorded now the machine will compare the recorded and the new sounds. Also speaker recognition is used to control of access to restrict the services more than that may be used be military operations to fire or bomb some areas by commanding the weapons by using merely words “shoot”
There are some challenges of speaker recognition which are encountered in speech recognition is about the exact unit to be recognized by the machine. On finding the boundaries of these units in the signal it is difficult to find what is the really voice of the user. Another problem is the variation of the sound due to some unavoidable circumstances e.g. influenza the sound vary and the machine deny to allow you to service to the really user.
To minimize the speaker recognition problem the following may be done
Ø Using the exact words required
Ø Maintain the voice pitch
Ø Be aware of the environment condition
The presentation aimed at giving the general view of how human beings are able to recognize someone’s voice by hearing in relatively to the machine how can identify two different voices. Speaker recognition is the process of automatically recognizing speaking basing on the recorded voice. The lesson was very pleasant as the clarification was quite easy to understand. The speaker recognition is of two kinds which are speaker verification and speaker identification
The system works under the principle of pitch variation, the sound pitch which is subjected into disparities is recorded baring into mind that the sound is in analogous signals. The interesting issue is how the analog speech is converted into digital signal then upon comparison of current speech and stores the previous speech.
The speaker recognition are used for sake of security purpose where the user has to do several trial of recording their voices so the machine is just taking the average of sound pitch and stores it. Once the user comes and wants to user the machine would repeat the same words which will be recorded now the machine will compare the recorded and the new sounds. Also speaker recognition is used to control of access to restrict the services more than that may be used be military operations to fire or bomb some areas by commanding the weapons by using merely words “shoot”
There are some challenges of speaker recognition which are encountered in speech recognition is about the exact unit to be recognized by the machine. On finding the boundaries of these units in the signal it is difficult to find what is the really voice of the user. Another problem is the variation of the sound due to some unavoidable circumstances e.g. influenza the sound vary and the machine deny to allow you to service to the really user.
To minimize the speaker recognition problem the following may be done
Ø Using the exact words required
Ø Maintain the voice pitch
Ø Be aware of the environment condition
Reference:
M.R Schroeder(1985) -Speech and speaker recognition.
http://www.speech.cs.cmu.edu/comp.speech/Section6/Q6.6.html
http://www.speech.cs.cmu.edu/comp.speech/Section6/Q6.6.html
No comments:
Post a Comment