Title:
VOICE SIGNAL PROCESSING METHOD AND DEVICE
Document Type and Number:
WIPO Patent Application WO/2016/183791
Kind Code:
A1
Abstract:
Disclosed in the embodiments of the present invention are a voice signal processing method and device. The method comprises: capturing, via an image capture apparatus, an image of a target speaker; determining a mouth region of the target speaker from the image and taking the region as a target voice source region; acquiring position information of the target voice source region relative to the image capture apparatus by point feature positioning; according to a pre-stored relative spatial position between the image capture apparatus and a microphone array, and the position information of the target voice source region relative to the image capture apparatus, determining a relative position of the target voice source region and the microphone array; upon receipt of the voice signal via the microphone array, according to the relative position of the target voice source region and the microphone array, enhancing the voice signal by utilizing a minimum variance distortionless response (MVDR) beamforming algorithm. The embodiments of the present invention can increase accuracy of voice source positioning and effectively improve a noise reduction effect of the voice signal.
Inventors:
ZHAO TIANYU (CN)
Application Number:
PCT/CN2015/079245
Publication Date:
November 24, 2016
Filing Date:
May 19, 2015
Export Citation:
Assignee:
HUAWEI TECH CO LTD (CN)
International Classes:
H04N5/232
Foreign References:
| CN104012074A | 2014-08-27 | |||
| CN101674410A | 2010-03-17 | |||
| JP2010233173A | 2010-10-14 |
Attorney, Agent or Firm:
GUANGZHOU SCIHEAD PATENT AGENT CO., LTD (CN)
广州三环专利代理有限公司 (CN)
广州三环专利代理有限公司 (CN)
Download PDF:
Previous Patent: USE OF PLANT CRUDE EXTRACTS OF PUERARIA PEDUNCULARIS FOR PREPARATION OF MOLLUSCICIDES
Next Patent: CURABLE AQUEOUS COMPOSITION
Next Patent: CURABLE AQUEOUS COMPOSITION