Access control application based on voiceprint recognition
Abstract
Voiceprint recognition is a type of speech recognition that can be commonly
referred to as speaker recognition. In recent years, with the artificial intelligence
gradually entering the public eye, the authentication method based on the neural
network has been further developed. Such as face recognition, iris recognition and
voiceprint recognition. These identity authentication methods gradually evolved from
the theoretical stage into practical applications.
The traditional access control requires the user to carry a certain key device such
as a key or an access control card. The user needs to directly contact the key device
with the authentication system, and the authentication system can only open the key
information. With the development of technology, identity authentication technology
continues to simplify the authentication process, optimize the authentication
efficiency, provide protection for information security, and increase convenience for
identity authentication. This topic uses voiceprint recognition to provide users with a
simple, intelligent authentication method that does not require carrying any object.
This topic reads the standard user speech, analyzes the characteristic parameters
spoken by the user, and establishes a probability and statistics model for it. When the
user to be identified wants to pass through the access control, the pre-set text is read
aloud, and the system collects the voice of the user to be identified, analyzes the
feature parameters, and then puts it into the standard user's model to calculate the
probability matching value. If this probability matching value satisfies a certain
condition, it indicates that the user to be identified is a standard user. After the
identification process is completed, the system sends the recognition log and the voice
of the person to be identified and the corresponding recognition situation to the server.
评论0