Processing

Please wait...

Settings

Settings

Goto Application

1. WO2019208137 - SOUND SOURCE SEPARATION DEVICE, METHOD THEREFOR, AND PROGRAM

Publication Number WO/2019/208137
Publication Date 31.10.2019
International Application No. PCT/JP2019/014817
International Filing Date 03.04.2019
IPC
H04R 3/00 2006.01
HELECTRICITY
04ELECTRIC COMMUNICATION TECHNIQUE
RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
3Circuits for transducers
G10L 21/0272 2013.01
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
21Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
02Speech enhancement, e.g. noise reduction or echo cancellation
0272Voice signal separating
G10L 21/0308 2013.01
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
21Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
02Speech enhancement, e.g. noise reduction or echo cancellation
0272Voice signal separating
0308characterised by the type of parameter measurement, e.g. correlation techniques, zero crossing techniques or predictive techniques
CPC
G10L 21/0272
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
21Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
02Speech enhancement, e.g. noise reduction or echo cancellation
0272Voice signal separating
G10L 21/0308
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
21Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
02Speech enhancement, e.g. noise reduction or echo cancellation
0272Voice signal separating
0308characterised by the type of parameter measurement, e.g. correlation techniques, zero crossing techniques or predictive techniques
H04R 3/00
HELECTRICITY
04ELECTRIC COMMUNICATION TECHNIQUE
RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
3Circuits for transducers ; , loudspeakers or microphones
Applicants
  • 日本電信電話株式会社 NIPPON TELEGRAPH AND TELEPHONE CORPORATION [JP]/[JP]
Inventors
  • 伊藤 弘章 ITO, Hiroaki
  • 小泉 悠馬 KOIZUMI, Yuma
  • 原田 登 HARADA, Noboru
Agents
  • 中尾 直樹 NAKAO, Naoki
  • 中村 幸雄 NAKAMURA, Yukio
  • 義村 宗洋 YOSHIMURA, Takahiro
Priority Data
2018-08309724.04.2018JP
Publication Language Japanese (JA)
Filing Language Japanese (JA)
Designated States
Title
(EN) SOUND SOURCE SEPARATION DEVICE, METHOD THEREFOR, AND PROGRAM
(FR) DISPOSITIF DE SÉPARATION DE SOURCES SONORES, PROCÉDÉ POUR SA MISE EN ŒUVRE ET PROGRAMME
(JA) 音源分離装置、その方法、およびプログラム
Abstract
(EN)
Provided is a sound source separation technology device having higher separation precision than conventional devices. The sound source separation device has: a diffuse noise removal unit that removes an estimated value for a diffuse noise acoustic signal from an observation signal and finds a removed signal; a filter design unit that obtains a filter by combining a probability distribution that models the removed signal and a probability distribution that models a transfer function; and a sound source separation unit that separates at least a first acoustic signal and an estimated value for a noise component including an interference noise acoustic signal, from the observation signal by using the filter.
(FR)
L'invention concerne un dispositif de technologie de séparation de source sonore ayant une précision de séparation plus élevée que les dispositifs classiques. Le dispositif de séparation de source sonore comprend : une unité d'élimination de bruit diffus qui élimine une valeur estimée pour un signal acoustique de bruit diffus à partir d'un signal d'observation et trouve un signal retiré; une unité de conception de filtre qui obtient un filtre en combinant une distribution de probabilité qui modélise le signal retiré et une distribution de probabilité qui modélise une fonction de transfert; et une unité de séparation de source sonore qui sépare au moins un premier signal acoustique et une valeur estimée pour une composante de bruit comprenant un signal acoustique de bruit d'interférence, à partir du signal d'observation à l'aide du filtre.
(JA)
従来よりも分離精度の高い音源分離技術装置を提供する。音源分離装置は、観測信号から拡散性雑音音響信号の推定値を除去し、除去済信号を求める拡散性雑音除去部と、除去済信号をモデル化した確率分布と、伝達関数をモデル化した確率分布と、を組み合わせることでフィルタを得るフィルタ設計部と、フィルタにより観測信号から、少なくとも第一音響信号と干渉性雑音音響信号を含む雑音成分の推定値とを分離する音源分離部と、を有する。
Also published as
Latest bibliographic data on file with the International Bureau