Parameters for a feature extractor and acoustic model of a speech recognition module are trained. An objective function is utilized to determine values for the feature extractor parameters and the acoustic model parameters.
PROBLEM TO BE SOLVED: To provide a system and a method for reducing noise in a pattern recognition signal by accurately estimating the noise. SOLUTION: The method and the system define a mapping random variable as a function including at least a clean signal random variable and a noise random variable. Then a model parameter is decided in which at least one aspect of a distribution of values of the mapping random variables is described. On the basis of a model parameter, an estimation value of a clean signal random variable is determined. In many embodiments of the present invention, the mapping random variable is a signal-to-noise ratio variable and the method and the system estimate a value of the signal-to-noise rate variable from the model parameter. COPYRIGHT: (C)2011,JPO&INPIT
Dong Yu (俞栋)合作论文数Tencent AI Lab23
Li Deng合作论文数Department of Electrical and Computer Engineering, University of Washington;Vatic Investments;RIDGE-LANE Limited Partners19
Jasha Droppo合作论文数Amazon16
Ye-Yi Wang合作论文数Microsoft Corporation12
Yun-Cheng Ju合作论文数Microsoft11
Xuedong Huang (黄学东)合作论文数9
Michael L. Seltzer合作论文数Meta8
Zhengyou Zhang (张正友)合作论文数Tencent7
Richard Stern合作论文数Department of Electrical and Computer Engineering, College of Engineering, Carnegie Mellon University;Language Technologies Institute, Carnegie Mellon University;Department of Computer Science, Carnegie Mellon University;School of Music, Carnegie Mellon University5
Zicheng Liu合作论文数AMD5
Milind Mahajan合作论文数Speech Research Group, Speech Research group, Microsoft5Peter Mau合作论文数5
Geoffrey Zweig合作论文数WorldQuant4
Yifan Gong合作论文数Microsoft4michael j sinclair合作论文数3
Kuansan Wang合作论文数3
Xiaodong He (何晓冬)合作论文数JD.COM Group;Department of Electrical & Computer Engineering, University of Washington3
Fu-Hua Liu合作论文数3
Xiaolong Li (李小龙)合作论文数Ant Group3
Amarnag Subramanya合作论文数Microsoft AI3leon c w wong合作论文数2christian huitema合作论文数2Jian Wu合作论文数2Hagai Attias合作论文数2
Mei-Yuh Hwang合作论文数Department of Electrical & Computer Engineering, University of Washington2
Jinyu Li合作论文数Microsoft2
Pedro J Moreno Mengibar合作论文数Humain2Issam Bazzi合作论文数2
Leo Tsz-On Lee (李子安)合作论文数Faculty of Health Sciences,University of Macau2kori marie quinn合作论文数1Daniel Tapias合作论文数1daniel rosenfeld合作论文数1Plumpe, M.D.合作论文数1
Jianfeng Gao (高剑峰)合作论文数Microsoft Research1
Xiao Li合作论文数DoorDash1
Dan Bohus合作论文数Microsoft Research1
Dinei Florencio合作论文数Microsoft Rersearch1
Alex Acero合作论文数Department of Electrical and Computer Engineering, University of Washington;Apple1
Philip A. Chou合作论文数Google1
Ross Cutler合作论文数Microsoft1
Hsiao-Wuen Hon (洪小文)合作论文数School of Artificial Intelligence, Xidian University;Microsoft Research Asia1
Brendan J. Frey合作论文数Department of Electrical and Computer Engineering, University of Toronto;Best Department of Medical Research, University of Toronto;Department of Computer Science, University of Toronto1
Henrique S. Malvar (Rico)合作论文数Microsoft Research1
Xing Fan合作论文数Triticeae Research Institute, Sichuan Agricultural University1
Eric Horvitz合作论文数Microsoft1Yanli Zheng合作论文数1
Ciprian Chelba合作论文数Google DeepMind;Google Research;Speech Recognition Research Community1
William (Bill) Buxton合作论文数Technical University of Eindhoven1
Po-Sen Huang合作论文数Google DeepMind1
