Back to results
US 12,711,960 B2Granted
Speaker recognition in the call center
AssigneePindrop Security, Inc.
InventorsElie Khoury, Matthew Garland
Life of this patent
Estimating the term…
Abstract
Utterances of at least two speakers in a speech signal may be distinguished and the associated speaker identified by use of diarization together with automatic speech recognition of identifying words and phrases commonly in the speech signal. The diarization process clusters turns of the conversation while recognized special form phrases and entity names identify the speakers. A trained probabilistic model deduces which entity name(s) correspond to the clusters.
Claims20
A computer-implemented method comprising:
- extracting, by the computer, an enrollment voiceprint for an enrolled speaker using a plurality of enrollment audio features of one or more enrollment speech samples, including a set of enrollment audio features for a keyword sequence occurring in each enrollment speech sample;
- extracting, by the computer, a test voiceprint for a test speaker using a plurality test audio features of a test speech sample;
- determining, by the computer, an audio similarity score based upon a difference between the enrollment voiceprint and the test voiceprint;
- and
- identifying, by the computer, a content match between the set of enrollment audio features for the keyword sequence and a set of test audio features of the plurality of test audio features of the test speech sample.
2 more elements
A system comprising:
- a non-transitory storage medium storing a plurality of computer program instructions;
- and
- a computer having at least one processor electrically coupled to the non-transitory storage medium and configured to execute the computer program instructions to:
- extract an enrollment voiceprint for an enrolled speaker using a plurality of enrollment audio features of one or more enrollment speech samples, including a set of enrollment audio features for a keyword sequence occurring in each enrollment speech sample;
- extract a test voiceprint for a test speaker using a plurality test audio features of a test speech sample;
- determine an audio similarity score based upon a difference between the enrollment voiceprint and the test voiceprint;
- and
- identify a content match between the set of enrollment audio features for the keyword sequence and a set of test audio features of the plurality of test audio features of the test speech sample.
5 more elements
Dependencies are inferred from the claim text.