Contents
System Characteristics
1. Memory:- The system has memory as the output depends on the current input as well on the input give at the
time of training.
2. Invertibility:- The system is non-invertible as there are only two outputs accept or deny for all the possible
inputs.
3. Causality:- The system is causal as the output depends only on the input signal and the key signal stored
during the training process.
4. Stability:- The system is stable as for any given input the output is bounded i.e. either its accept or deny.
5. Time invariance:- The system is time invariant as on keeping the input signal constant, the output is also
unchanged irrespective of the time.
6. Linearity:- The system is non-linear as on addition of two input signals does not give an output which is the
sum of their respective outputs. Homogeneity also doesn’t hold as the output doesn’t scale for a scaled input.
USE OF APPLICATION IN REAL LIFE:
a) Speaker verification is used for secure access to services via the telephone, including home shopping, home banking and telecom services.b) In a forensic context, speaker verification can be deployed in the processing of telephone taps, either to identify a talking suspect, or track the time intervals where a suspect is talking in a lengthy telephone tap . c) Another emerging market is secure access to information which can be obtained through the internet.
Multimodality is a hot issue, here as well as elsewhere: there is a need for a better understanding of the ways in which natural speech can be combined with other media for both input and output.
d) When speech of a visitor is available and security is an issue, speaker verification can be deployed to verify the visitor's identity .
e) One other important application is probation monitoring, where speaker verification is applied to monitor the whereabouts of persons who are not allowed to travel freely, e.g. rehabilitating prisoners.
ACKNOWLEDGMENT:
We would like to thank Prof. V. M. Gadre and the course associates of EE210 for their valuable time and guidance.REFERENCES
1. ROBERT D. RODMAN : Computer Speech Technology2. DOUGLAS O’ SHAUGHNESSY : Speech Communication- Human and Machines
3. SADAVKI FURUI , MOHAN SONDHI : Advances in Speech Signal Processing
4. 1999 IEEE International Conference on Acoustic, Speech and Signal Processing. Volume I.
