发明名称 APPARATUS AND METHOD FOR VOICE SIGNAL PROCESSING
摘要 PROBLEM TO BE SOLVED: To provide an apparatus and a method for voice signal processing capable of compensating a mismatch by presuming a mismatch in sampling frequency in respective voice signals at high accuracy and even in a long recording.SOLUTION: A voice signal processing apparatus synchronizes a subject channel signal with a reference channel signal when a recording starting time difference exists in between the subject channel signal and the reference channel signal and a mismatch exists in sampling frequencies between A/D converters of respective recording equipment. While deriving a reference channel signal in a short time Fourier transformation expression by carrying out a constant frame shift for the reference channel signal, the voice signal processing apparatus derives a subject channel signal in the short time Fourier transformation expression by carrying out a small number sample shift by an integer sample shift and a phase compensation method in frequency range for the subject channel signal so that frame centers for each correspond to each other based on a mismatch between the recording time difference and a sampling frequencies .
申请公布号 JP2014174393(A) 申请公布日期 2014.09.22
申请号 JP20130048084 申请日期 2013.03.11
申请人 RESEARCH ORGANIZATION OF INFORMATION & SYSTEMS ; UNIV OF TSUKUBA 发明人 ONO JUNKI ; MIYABE SHIGEKI ; MAKINO SHOJI
分类号 G10L21/0272 主分类号 G10L21/0272
代理机构 代理人
主权项
地址