Objective speech quality is difficult to be measured without the input reference speech.Mapping methods using data mining are investigated and designed to improve the output-based speech quality assessment algorithm.T...Objective speech quality is difficult to be measured without the input reference speech.Mapping methods using data mining are investigated and designed to improve the output-based speech quality assessment algorithm.The degraded speech is firstly separated into three classes(unvoiced,voiced and silence),and then the consistency measurement between the degraded speech signal and the pre-trained reference model for each class is calculated and mapped to an objective speech quality score using data mining.Fuzzy Gaussian mixture model(GMM)is used to generate the artificial reference model trained on perceptual linear predictive(PLP)features.The mean opinion score(MOS)mapping methods including multivariate non-linear regression(MNLR),fuzzy neural network(FNN)and support vector regression(SVR)are designed and compared with the standard ITU-T P.563 method.Experimental results show that the assessment methods with data mining perform better than ITU-T P.563.Moreover,FNN and SVR are more efficient than MNLR,and FNN performs best with 14.50% increase in the correlation coefficient and 32.76% decrease in the root-mean-square MOS error.展开更多
A differential optical absorption spectroscopy (DOAS) method is presented and used for the air pollutant detection. The novel measurement frame employs a Cassegrain telescope and a combined fiber bundle. The emitting ...A differential optical absorption spectroscopy (DOAS) method is presented and used for the air pollutant detection. The novel measurement frame employs a Cassegrain telescope and a combined fiber bundle. The emitting and receiving fibers are bundled together at one port. The common port is placed at the focus of the Cassegrain telescope. The total length of the prototype is reduced to about 1/2 of the traditional one. Air pollutants of SO2, NO2 and O3 are monitored by the prototype, and the concentrations are inverted. Results show that the correlation coefficients are 0.9490, 0.9614 and 0.9301, respectively. And there is a good consistency between the measured results and the reference data.展开更多
基金Projects(61001188,1161140319)supported by the National Natural Science Foundation of ChinaProject(2012ZX03001034)supported by the National Science and Technology Major ProjectProject(YETP1202)supported by Beijing Higher Education Young Elite Teacher Project,China
文摘Objective speech quality is difficult to be measured without the input reference speech.Mapping methods using data mining are investigated and designed to improve the output-based speech quality assessment algorithm.The degraded speech is firstly separated into three classes(unvoiced,voiced and silence),and then the consistency measurement between the degraded speech signal and the pre-trained reference model for each class is calculated and mapped to an objective speech quality score using data mining.Fuzzy Gaussian mixture model(GMM)is used to generate the artificial reference model trained on perceptual linear predictive(PLP)features.The mean opinion score(MOS)mapping methods including multivariate non-linear regression(MNLR),fuzzy neural network(FNN)and support vector regression(SVR)are designed and compared with the standard ITU-T P.563 method.Experimental results show that the assessment methods with data mining perform better than ITU-T P.563.Moreover,FNN and SVR are more efficient than MNLR,and FNN performs best with 14.50% increase in the correlation coefficient and 32.76% decrease in the root-mean-square MOS error.
基金the National Natural Science Foundation of China (No.60938002)the National SME Technology Innovation Fund of China (No.10C26211203591)the University Science and Technology Development Fund of Tianjin in China (No.20120522)
文摘A differential optical absorption spectroscopy (DOAS) method is presented and used for the air pollutant detection. The novel measurement frame employs a Cassegrain telescope and a combined fiber bundle. The emitting and receiving fibers are bundled together at one port. The common port is placed at the focus of the Cassegrain telescope. The total length of the prototype is reduced to about 1/2 of the traditional one. Air pollutants of SO2, NO2 and O3 are monitored by the prototype, and the concentrations are inverted. Results show that the correlation coefficients are 0.9490, 0.9614 and 0.9301, respectively. And there is a good consistency between the measured results and the reference data.
