System and method for performing voice biometrics analysis

1. A method for audio analysis, performed by an at least one processor, the method comprising:

a. sending part of a stream of audio to a voice processor, wherein the voice processor is configured to extract features of the audio of the part of a stream of audio;

b. obtaining the extracted features from the voice processor;

c. calculating, based on the extracted features, a quality grade of the part of a stream of audio; and

d. sending the extracted features to an at least one voice biometrics engine if the quality grade is above a threshold;

wherein each of the at least one voice biometrics engines is configured to generate a voiceprint of the part of a stream of audio, based on the extracted features of the part of a stream of audio.

2. The method of claim 1 , wherein each of the at least one voice biometrics engines is further configured to perform at least one of:

authenticate a speaker by comparing a voiceprint of the part of a stream of audio to a previously stored voiceprint of the same speaker,

detect fraudsters by comparing the voiceprint of the part of a stream of audio to previously stored voiceprints of known fraudsters; and

enrich the previously stored voiceprint of the speaker with the voiceprint of the part of a stream of audio.

3. The method of claim 1 , comprising:

obtaining the voiceprint;

sending the voiceprint to an at least one voiceprint analyzer, wherein each of the at least one voiceprint analyzers is configured to perform at least one of:

authenticate a speaker by comparing a voiceprint of the part of a stream of audio to a previously stored voiceprint of the same speaker,

detect fraudsters by comparing the voiceprint of the part of a stream of audio to previously stored voiceprints of known fraudsters; and

enrich the previously stored voiceprint of the speaker with the voiceprint of the first part of a stream of audio.

4. The method of claim 3 , comprising:

obtaining the voiceprint by the at least one voiceprint analyzer; and

performing, by each of the at least one voiceprint analyzer, at least one of: authenticating the speaker, detecting the fraudsters and enriching the previously stored voiceprint, in parallel.

5. The method of claim 1 , comprising:

e. storing the extracted features in a repository if the quality grade is below the threshold;

f. upon receiving a different part of a stream of audio from a same speaker as the part of a stream of audio, retrieving the voice features of the part of a stream of audio; and

g. sending the different part of a stream of audio to voice processor together with the retrieved voice features of the part of a stream of audio, wherein the voice processor is configured to modify the extracted features based on the different part of a stream of audio.

6. The method of claim 5 , comprising:

h. obtaining the modified extracted features from the voice processor;

i. calculating, based on the modified extracted features, an updated quality grade;

j. sending the modified extracted features to the at least one voice biometrics engine if the updated quality grade is above the threshold;

k. storing the modified extracted features in the repository if the updated quality grade is below the threshold; and

l. repeating operations f-k until a session with the speaker is terminated.

7. The method of claim 6 , wherein the updated quality grade equals a total amount of speech in the part of a stream of audio and the different part of a stream of audio.

8. The method of claim 1 , wherein the at least one processor is part of a cloud-based computing system, and wherein the voice processor and the voice biometrics engine are micro-services of the cloud-based computing system.

9. The method of claim 1 , wherein the quality grade equals an amount of speech in the part of a stream of audio.

10. A system for audio analysis, the system comprising:

a plurality of processors in a cloud-based computerized system, wherein the processors are configured to execute a plurality of micro-service types in parallel, wherein an authentication (RTA) manager micro-service is configured to:

a. send a part of a stream of audio to a voice processor micro-service, wherein the voice processor micro-service is configured to extract features of the audio of the part of a stream of audio;

b. obtain the extracted features from the voice processor micro-service;

c. calculate, based on the extracted features, a quality grade of the part of a stream of audio; and

d. send the extracted features to an at least one voice biometrics engine micro-service if the quality grade is above a threshold;

wherein each of the at least one voice biometrics engine micro-services is configured to generate a voiceprint of the part of a stream of audio, based on the extracted features of the part of a stream of audio.

11. The system of claim 10 , wherein each of the at least one voice biometrics engine micro-services is further configured to perform at least one of:

authenticate a speaker by comparing a voiceprint of the part of a stream of audio to a previously stored voiceprint of the same speaker,

detect fraudsters by comparing the voiceprint of the part of a stream of audio to previously stored voiceprints of known fraudsters; and

enrich the previously stored voiceprint of the speaker with the voiceprint of the part of a stream of audio.

12. The system of claim 10 , wherein the RTA manager micro-service is configured to:

obtain the voiceprint;

send the voiceprint to an at least one voiceprint analyzer micro-service, wherein each of the at least one voiceprint analyzer micro-services is configured to perform at least one of:

authenticate a speaker by comparing a voiceprint of the part of a stream of audio to a previously stored voiceprint of the same speaker,

detect fraudsters by comparing the voiceprint of the part of a stream of audio to previously stored voiceprints of known fraudsters; and

enrich the previously stored voiceprint of the speaker with the voiceprint of the part of a stream of audio.

13. The system of claim 10 , wherein the RTA manager micro-service is configured to:

e. store the extracted features in a repository if the quality grade is below the threshold;

f. upon receiving a different part of a stream of audio from a same speaker as the part of a stream of audio, retrieve the voice features of the part of a stream of audio; and

g. send the different part of a stream of audio to voice processor micro-service together with the retrieved voice features of the part of a stream of audio, wherein the voice processor micro-service is configured to modify the extracted features based on the different part of a stream of audio.

14. The system of claim 13 , wherein the RTA manager micro-service is configured to:

h. obtain the modified extracted features from the voice processor micro-service;

i. calculate, based on the modified extracted features, an updated quality grade;

j. send the modified extracted features to the at least one voice biometrics engine micro-service if the updated quality grade is above the threshold;

k. store the modified extracted features in the repository if the updated quality grade is below the threshold; and

l. repeat operations f-k until a session with the speaker is terminated.

15. The system of claim 14 , wherein the updated quality grade equals a total amount of speech in the part of a stream of audio and the different part of a stream of audio.

16. The system of claim 10 , wherein the quality grade equals an amount of speech in the part of a stream of audio.

17. A method comprising:

transmitting by a processor a part of a stream of audio to a process configured to calculate features of the part of a stream of audio;

determining by the processor, based on the features, a rating for the part of a stream of audio; and

transmitting the extracted features to an at least one engine to create a voiceprint from the features if the rating is above a threshold;

wherein each of the at least one voice biometrics engine micro-services is configured to generate a voiceprint of the part of a stream of audio, based on the extracted features of the part of a stream of audio.