Study on robust speech by component movement on time frequency plane and construction of speech database for researchers

2016 Fiscal Year Final Research Report

• Project/Area Number: 26330188
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Perceptual information processing
• Research Institution: The University of Electro-Communications
• Principal Investigator: Takahashi Kota 電気通信大学, 大学院情報理工学研究科, 准教授 (10188005)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 話速変換 / 音声データベース / 時間周波数平面

Outline of Final Research Achievements

For the method of enhancing the robustness of speech against the disturbance sound, we proposed a method to solve separately improvement of hearing of vowel part and hearing improvement of consonant part by combining linear prediction analysis and modulation spectrum analysis method. Experiments showed the effectiveness of the proposed method. We also showed that it is effective to improve listening by expanding the acoustic signal on the time frequency plane and moving its components. In addition, in this study, we aimed to expand our original speech database which was created and made public for audition test of speech. During that period, we enhanced the "one character difference" speech set and the "pseudo car navigation" speech set, and made it available for researchers to use easily.

Free Research Field

ディジタル信号処理