Title:
SPEECH PROCESSING DEVICE, SPEECH PROCESSING METHOD, AND SPEECH PROCESSING PROGRAM
Document Type and Number:
WIPO Patent Application WO/2024/023946
Kind Code:
A1
Abstract:
A first training device (10) calculates a first loss function that becomes smaller as a vector into which a model has quantized the feature amount of speech becomes closer to a context representation which the model has acquired from the feature amount, and a second loss function that becomes smaller as an accuracy with which the model identifies meta information of the speech on the basis of the context representation becomes higher. The first training device (10) updates the parameter of the model such that the first loss function becomes smaller and the second loss function becomes larger.
Inventors:
TANAKA TOMOHIRO (JP)
MASUMURA RYO (JP)
MASUMURA RYO (JP)
Application Number:
PCT/JP2022/028843
Publication Date:
February 01, 2024
Filing Date:
July 26, 2022
Export Citation:
Assignee:
NIPPON TELEGRAPH & TELEPHONE (JP)
International Classes:
G10L15/16; G10L15/00
Domestic Patent References:
WO2022044243A1 | 2022-03-03 |
Foreign References:
CN113327595A | 2021-08-31 | |||
CN112951213A | 2021-06-11 |
Attorney, Agent or Firm:
SAKAI INTERNATIONAL PATENT OFFICE (JP)
Download PDF:
Previous Patent: RELEASE FILM AND METHOD FOR MANUFACTURING SEMICONDUCTOR PACKAGE
Next Patent: LEARNING DEVICE, LEARNING METHOD, AND LEARNING PROGRAM
Next Patent: LEARNING DEVICE, LEARNING METHOD, AND LEARNING PROGRAM