Presentation 2017-12-14
Harmonic Structure Detection in Speech Separation Using Modified DFT Pair Based on ASA
Motohiro Ichikawa, Isao Nakanishi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Humans have the ability of cocktail party effect to be able to recognized the target voice from the various conversations. Auditory scene analysis (ASA) is a framework that turns its ability into model. In this study, we propose speech separation by decomposing mixed sounds with single microphone input sequentially into frequency signals using Modified DFT pairs, extracting the ASA’s four features, and regrouping frequency signals. In this paper, we improve the detection part of the harmonic structure that is the core in the speech separation system constructed by the previous study, and improve the separation accuracy.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Speech Separation / Auditory Scene Analysis / Computational Auditory Scene Analysis / Sequential Implementation / Harmonic Structure
Paper # SIS2017-34
Date of Issue 2017-12-07 (SIS)

Conference Information
Committee SIS
Conference Date 2017/12/14(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Tottori Prefectural Center for Lifelong Learning
Topics (in Japanese) (See Japanese page)
Topics (in English) System on Silicon, Short Range Wireless Systems, etc.
Chair Hirokazu Tanaka(Hiroshima City Univ.)
Vice Chair Takayuki Nakachi(NTT) / Noriaki Suetake(Yamaguchi Univ.)
Secretary Takayuki Nakachi(Kanagawa Inst. of Tech.) / Noriaki Suetake(Kyushu Inst. of Tech.)
Assistant Masaaki Fujiyoshi(Tokyo Metropolitan Univ.) / Takanori Koga(NIT, Tokuyama Col.)

Paper Information
Registration To Technical Committee on Smart Info-Media Systems
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Harmonic Structure Detection in Speech Separation Using Modified DFT Pair Based on ASA
Sub Title (in English)
Keyword(1) Speech Separation
Keyword(2) Auditory Scene Analysis
Keyword(3) Computational Auditory Scene Analysis
Keyword(4) Sequential Implementation
Keyword(5) Harmonic Structure
1st Author's Name Motohiro Ichikawa
1st Author's Affiliation Tottori University(Tottori Univ)
2nd Author's Name Isao Nakanishi
2nd Author's Affiliation Tottori University(Tottori Univ)
Date 2017-12-14
Paper # SIS2017-34
Volume (vol) vol.117
Number (no) SIS-349
Page pp.pp.5-9(SIS),
#Pages 5
Date of Issue 2017-12-07 (SIS)