Voiceprint recognition method, device, storage medium and background server

A voiceprint recognition device, comprising:

• a client configured to collect a test voice of a user and send a voice recognition request to the background server, the voice recognition request comprises a user ID and the test voice;
• a background server configured to receive the voice recognition request and determine the voice recognition request to be processed with a message queue and an asynchronous mechanism;
• a background server configured to acquire a target voiceprint feature which corresponds to the user ID of the voice recognition request to be processed, and acquire a test voiceprint feature which corresponds to the test voice of the voice recognition request to be processed;
• a background server configured to judge, according to the target voiceprint feature and the test voiceprint feature, whether the target voiceprint feature and the test voiceprint feature correspond to the same user, and output the result of the judging to the client;
• and
• a client configured to receive and display the result of the judging;
• wherein the background server comprises:
• a feature dimensionality reducing unit configured to reduce the dimensionality of the target voiceprint feature and the test voiceprint feature respectively with a PLDA algorithm, to acquire the target dimensionality reduction value and the test dimensionality reduction value;
• a cosine measurement processing unit configured to perform cosine measurement on the target dimensionality reduction value and the test dimensionality reduction value using a cosine measurement function to acquire a cosine measurement value;
• and
• a user recognition judging unit configured to judge whether the cosine measurement value is greater than a similarity threshold;
• wherein if the cosine measurement value is greater than a similarity threshold, the target voiceprint feature and the test voiceprint feature correspond to the same user;
• and if the cosine measurement value is not greater than a similarity threshold, the target voiceprint feature and the test voiceprint feature do not correspond to the same user.

A background server comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of:

• collecting, by a client, a test voice of a user, and sending a voice recognition request to the background server, the voice recognition request comprises a user ID and the test voice;
• receiving, by the background server, the voice recognition request, and determining the voice recognition request to be processed with a message queue and an asynchronous mechanism;
• acquiring, by the background server, a target voiceprint feature which corresponds to the user ID of the voice recognition request to be processed, and acquiring a test voiceprint feature which corresponds to the test voice of the voice recognition request to be processed;
• judging, by the background server according to the target voiceprint feature and the test voiceprint feature, whether the target voiceprint feature and the test voiceprint feature correspond to the same user, and outputting the result of the judging to the client;
• and
• receiving and displaying, by the client, the result of the judging;
• wherein it is judged according to the target voiceprint feature and the test voiceprint feature whether the target voiceprint feature and the test voiceprint feature correspond to the same user, comprising:
• applying dimensionality reduction to the target voiceprint feature and the test voiceprint feature respectively with a PLDA algorithm, to acquire the target dimensionality reduction value and the test dimensionality reduction value;
• performing cosine measurement on the target dimensionality reduction value and the test dimensionality reduction value using a cosine measurement function to acquire a cosine measurement value;
• and
• judging whether the cosine measurement value is greater than a similarity threshold;
• wherein if the cosine measurement value is greater than a similarity threshold, the target voiceprint feature and the test voiceprint feature correspond to the same user;
• and if the cosine measurement value is not greater than a similarity threshold, the target voiceprint feature and the test voiceprint feature do not correspond to the same user.