Biometrics platform provider
View Patent ↗Systems and methods for analyzing digital recordings of the human voice in order to find characteristics unique to an individual. A biometrics engine may use an analytics service in a contact center to supply audio streams based on configured rules and providers for biometric detection. The analytics service may provide call audio data and attributes to connected engines based on a provider-set of selection rules. The connected providers send call audio data and attributes through the analytics service. The engines are notified when a new call is available for processing and can then retrieve chunks of audio data and call attributes by polling an analytics service interface. A mathematical model of the human vocal tract in the call audio data is created and/or matched against existing models. The result is analogous to a fingerprint, i.e., a pattern unique to an individual to within some level of probability.
1. A method for providing a biometrics platform, comprising:
providing at least one provider that receives and processes audio data associated with calls;
providing an analytics service at a recorder that receives the audio data from the at least one provider and provides metadata and the audio data to a biometrics engine and an enrollment engine in accordance with rules associated with the at least one provider;
providing a person database that electronically stores voiceprints that are each uniquely associated with a person; and
registering the biometrics engine with the at least one provider in accordance with the rules to forward the audio data to a detection library to make a comparison of the audio data and metadata with electronically stored voiceprints and electronically stored metadata in the person database, receive comparison results from the detection library, and to call the analytics service to tag a call based on the comparison results of the audio data to the voiceprints.
2. The method of claim 1 , further comprising:
interleaving audio chunks from multiple calls by the biometrics engine into a single stream;
sending the single stream over a single network socket to a streaming daemon; and
communicating the single stream from the streaming daemon to the detection library.
3. The method of claim 1 , wherein each voiceprint in the person database has weighted value associated with it that is indicative of a quality, authenticity and uniqueness, and wherein the weighted value is used by the biometrics platform to decide future actions to be taken in accordance with the comparison.
4. The method of claim 1 , further comprising providing an enrollment engine that performs auto enrollment by registering an interest in audio data for a specific provider.
5. The method of claim 1 , further comprising electronically storing person and attribute details with each voiceprint in the person database.
6. The method of claim 5 , further comprising electronically storing at least one of name, telephone number(s), gender, language(s) spoken and specific identifying information as the attribute details.
7. The method of claim 1 , further comprising electronically storing information specific to each voiceprint, the information including how many calls were used to create the voiceprint and how much audio was used to create the voiceprint.
8. The method of claim 1 , further comprising exposing interfaces of the detection library using application programming interfaces (APIs) to provide for querying, creating, updating and deleting voiceprints in the person database.
9. The method of claim 1 , the comparison further comprising;
normalizing a voiceprint against a set of existing voiceprints;
testing the voiceprint against a set of existing voiceprints; and
determining a Likelihood Ratio (LR), which is calculated by a logarithmic transformation to result in a value between −16 and +16.
10. The method of claim 1 , further comprising:
applying selection rules to determine which calls flow through the analytics service;
applying analytics rules to determine if the biometrics engine or the enrolment engine should process a call.
11. The method of claim 10 , wherein the analytics rules specify a Keyword Spotting Analytics Rule that is a string that identifies the words to spot and details of a predetermined combination of words that are considered to be a match.
12. The method of claim 11 , wherein the analytics rules provide a Boolean result.
13. The method of claim 10 , further comprising sending, by each provider, an audio stream to the analytics service where it is buffered until a particular selection rule is matched.
14. The method of claim 13 , further comprising:
starting to tag the call when the particular selection rule is matched; and
forwarding tagged calls to either the biometrics engine or the enrolment engine.
15. The method of claim 1 , further comprising performing a subsequent action based on tagging of the call.
16. The method of claim 15 , further comprising performing constructing a metadata dictionary for each call that contains entries for every available unique metadata tag.