Patent · US Active

Blind diarization of recorded calls with arbitrary number of speakers

US10109280B2 · kind B2 · utility

8Cited by

52References

19Claims

0Family size

Assignee

VERINT SYSTEMS INC. · IL

Inventors

Oana Sidi · Ramat HaSharon, IL
Ron Wein · Ramat HaSharon, IL

Key dates

Filing date	Dec 12, 2017
Grant date	Oct 23, 2018
Priority date	—
Expiry date	Dec 12, 2037

Classification

Technology area (CPC H)Electricity
CPC primaryH04M2203/303
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

In a method of diarization of audio data, audio data is segmented into a plurality of utterances. Each utterance is represented as an utterance model representative of a plurality of feature vectors. The utterance models are clustered. A plurality of speaker models are constructed from the clustered utterance models. A hidden Markov model is constructed of the plurality of speaker models. A sequence of identified speaker models is decoded.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.