Title:
VOICE EXTRACTION DEVICE, VOICE EXTRACTION METHOD AND VOICE EXTRACTION PROGRAM
Document Type and Number:
Japanese Patent JP2019049685
Kind Code:
A
Abstract:
To improve accuracy of voice recognition by extracting a voice properly.SOLUTION: A voice extraction device according to the present application includes a formation unit, an acquisition unit, a generation unit, and a selection unit. The formation unit forms directivity in advance by beam forming processing for each microphone in a microphone array in which a plurality of channels are formed by having a plurality of microphones. The acquisition unit acquires an observation signal which is a signal of a voice received by each channel. The generation unit generates, for each channel, frequency distribution of the amplitude of an emphasis signal generated by emphasizing the observation signal of each channel according to the directivity of each microphone formed by the formation unit. The selection unit selects a channel corresponding to a voice signal used for voice recognition among each of the channels on the basis of the frequency distribution corresponding to each channel generated by the generation unit.SELECTED DRAWING: Figure 4
Inventors:
OMACHI MOTOKI
Application Number:
JP2018023284A
Publication Date:
March 28, 2019
Filing Date:
February 13, 2018
Export Citation:
Assignee:
YAHOO JAPAN CORP
International Classes:
G10L15/20; G10L15/00; G10L21/0208; G10L21/0224; H04R1/40; H04R3/12
Attorney, Agent or Firm:
Sakai International Patent Office