Title:
ACOUSTIC PROCESSING METHOD AND ACOUSTIC PROCESSING SYSTEM
Document Type and Number:
WIPO Patent Application WO/2020/095951
Kind Code:
A1
Abstract:
This acoustic processing system is provided with: a learning processing unit that carries out additional learning of a learned synthesis model for generating, from condition data which represents a sound-generation condition, feature data which represents a feature of a sound generated under the sound-generation condition, by using first condition data representing the sound-generation condition identified from a sound signal and first feature data representing a feature of a sound represented by the sound signal; an instruction reception unit that receives an instruction for changing the sound-generation condition about the sound signal; and a synthesis processing unit that generates second feature data by inputting second condition data which represents the changed sound-generation condition into the synthesis model after the additional learning.
Inventors:
DAIDO RYUNOSUKE (JP)
Application Number:
PCT/JP2019/043511
Publication Date:
May 14, 2020
Filing Date:
November 06, 2019
Export Citation:
Assignee:
YAMAHA CORP (JP)
International Classes:
G10H1/00; G10L13/00; G10L13/033
Foreign References:
| JP2017107228A | 2017-06-15 | |||
| JP2016020972A | 2016-02-04 |
Other References:
NAKANO, TOMOYASU; GOTO, MASATAKA: "VocaListener: A Singing Synthesis System by Mimicking Pitch and Dynamics of User's Singing", TRANSACTIONS OF INFORMATION PROCESSING SOCIETY OF JAPAN, vol. 52, no. 12, 15 December 2011 (2011-12-15), pages 3853 - 3867, XP009527468, ISSN: 1882-7799
See also references of EP 3879521A4
See also references of EP 3879521A4
Attorney, Agent or Firm:
OHBAYASHI, Akira et al. (JP)
Download PDF:
Previous Patent: INFORMATION PROCESSING METHOD AND INFORMATION PROCESSING SYSTEM
Next Patent: ABSORBENT ARTICLE
Next Patent: ABSORBENT ARTICLE