Title:
VOICE SIGNAL-BASED METHOD AND APPARATUS FOR RECOGNIZING STRESS USING ADVERSARIAL TRAINING WITH SPEAKER INFORMATION
Document Type and Number:
WIPO Patent Application WO/2022/114347
Kind Code:
A1
Abstract:
The present embodiments provide apparatus and method for recognizing stress in which accuracy for recognizing stress is improved, the method using domain adversarial training with speaker information obtained during a training period to eliminate speaker-dependent traits from the voice signal, and independently training the characteristic vector associated with psychological stress with speaker information.
Inventors:
KANG HONG GOO (KR)
HAN HYE WON (KR)
SHIN HYEON KYEONG (KR)
BYUN KYUNG GUEN (KR)
HAN HYE WON (KR)
SHIN HYEON KYEONG (KR)
BYUN KYUNG GUEN (KR)
Application Number:
PCT/KR2020/017789
Publication Date:
June 02, 2022
Filing Date:
December 07, 2020
Export Citation:
Assignee:
UIF UNIV INDUSTRY FOUNDATION YONSEI UNIV (KR)
International Classes:
G10L25/63; A61B5/00; A61B5/16; G10L15/02; G10L17/02; G10L17/04
Foreign References:
| KR20200114705A | 2020-10-07 | |||
| KR20190135916A | 2019-12-09 | |||
| US20200364539A1 | 2020-11-19 | |||
| CN109308318A | 2019-02-05 |
Other References:
ABDELWAHAB MOHAMMED, BUSSO CARLOS: "Domain Adversarial for Acoustic Emotion Recognition", IEEE/ACM TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, IEEE, USA, vol. 26, no. 12, 1 December 2018 (2018-12-01), USA, pages 2423 - 2435, XP055933927, ISSN: 2329-9290, DOI: 10.1109/TASLP.2018.2867099
Attorney, Agent or Firm:
WOOIN PATENT & LAW FIRM (KR)
Download PDF: