Back to results
US 12,664,989 B2Granted
Speaker recognition device and operating method thereof
AssigneeLG Electronics Inc.
InventorSang Seok Lee
Life of this patent
Estimating the term…
Abstract
An object of the present disclosure is to ensure speaker recognition performance even if voice uttered by a voice input device different from a voice input device of voice uttered in a speaker registration process.
Claims9
A speaker recognition device comprising:
- a communication interface configured to communicate with an electronic device;
- a memory configured to store a speaker profile including a first voice feature set including voice features of a first voice signal and a first voice input channel matched to the first voice signal;
- and
- a processor configured to receive a second voice signal corresponding to a voice command uttered by a speaker and information on a second voice input channel indicating a device obtaining the voice command, from an electronic device,
- obtain a similarity between the first voice feature set and a second voice feature set including voice features of the second voice signal, based on that the second voice input channel is not the first voice input channel stored in the memory,
- transmit a notification indicating that the speaker is identified, to the electronic device, based on the obtained similarity being equal to or greater than a first similarity,
- transmit a notification for checking that the speaker is matched to a pre-registered speaker, to the electronic device, based on the similarity being less than the first similarity and equal to or greater than a second similarity less than the first similarity, and
- transmit a notification for registering a new speaker to the electronic device, based on the similarity being less than the second similarity.
5 more elements
A speaker recognition device comprising:
- a communication interface configured to communicate with an electronic device;
- a memory configured to store a speaker profile including a first voice feature set including voice features of a first voice signal and a first voice input channel matched to the first voice signal;
- and
- a processor configured to receive a second voice signal corresponding to a voice command uttered by a speaker and information on a second voice input channel indicating a device obtaining the voice command, from an electronic device,
- obtain a similarity between the first voice feature set and a second voice feature set including voice features of the second voice signal, based on the second voice input channel being different from the first voice input channel stored in the memory,
- transmit a notification indicating that the speaker is identified, to the electronic device, based on the obtained similarity being equal to or greater than a first similarity, and
- transmit a notification for registering the second voice input channel to the electronic device, based on the similarity being less than the first similarity and equal to or greater than a second similarity less than the first similarity.
4 more elements
An operating method of a speaker recognition device, the method comprising:
- storing a speaker profile including a first voice feature set including voice features of a first voice signal and a first voice input channel matched to the first voice signal;
- receiving a second voice signal corresponding to a voice command uttered by a speaker and information on a second voice input channel indicating a device obtaining the voice command, from an electronic device;
- obtaining a similarity between the first voice feature set and a second voice feature set including voice features of the second voice signal, based on that the second voice input channel is not the stored first voice input channel;
- and
- transmitting a notification indicating that the speaker is identified, to the electronic device, based on the obtained similarity being equal to or greater than a first similarity,
- transmitting a notification for checking that the speaker is matched to a pre-registered speaker, to the electronic device, based on the similarity being less than the first similarity and equal to or greater than a second similarity less than the first similarity, and
- transmitting a notification for registering a new speaker to the electronic device, based on the similarity being less than the second similarity.
4 more elements
Dependencies are inferred from the claim text.