DEEP NEURAL NETWORK-BASED METHOD AND APPARATUS FOR COMBINED NOISE AND ECHO REMOVAL

Title:

DEEP NEURAL NETWORK-BASED METHOD AND APPARATUS FOR COMBINED NOISE AND ECHO REMOVAL

Document Type and Number:

WIPO Patent Application WO/2018/190547

Kind Code:

A1

Abstract:

Disclosed is a deep neural network-based method and apparatus for combined noise and echo removal. The deep neural network-based method for combined noise and echo removal according to one embodiment of the present invention may comprise the steps of: extracting a feature vector from an audio signal that includes noise and echo; and acquiring a final audio signal from which both noise and echo have been removed, by using a combined noise and echo removal gain estimated by means of the feature vector and deep neural network (DNN) learning.

Inventors:

CHANG JOON-HYUK (KR)
SEO HYEJI (KR)

Application Number:

PCT/KR2018/003840

Publication Date:

October 18, 2018

Filing Date:

April 02, 2018

Export Citation:

Click for automatic bibliography generation Help

Assignee:

UNIV HANYANG IND UNIV COOP FOUND (KR)

International Classes:

G10L21/02; G10L19/028; G10L19/03; G10L19/08; G10L25/30

Foreign References:

US20170061978A1	2017-03-02
JP2013037174A	2013-02-21
US20140257803A1	2014-09-11
US20110238416A1	2011-09-29
JP2014157323A	2014-08-28

Attorney, Agent or Firm:

YANG, Sungbo (KR)

Download PDF:

View/Download PDF PDF Help

Previous Patent: ELECTRONIC APPARATUS AND CONTROL METHOD THEREOF

Next Patent: AIR CONDITIONER FOR VEHICLE