Recognition of non-specific two-word Chinese vocabulary by applying Fourier transform twice to the narrow-band spectrogram
Di Pan, Shili Liang, Ying Wei, Tingfa Xu, Shuangwei Wang · 2016
This paper illustrates a method to recognize nonspecific two-word Chinese vocabulary by analyzing speech signals using a narrow-band spectrogram after Fourier transform is applied to it twice. Binary width zoning line projection is carried out in the frequency domain. The projection value is treated as the characteristic value of semantic recognition feature and the support vector machine (SVM) is considered as the classifier for recognizing the semantics of non-specific two-word Chinese vocabulary. A total of 1000 voice samples were used in the simulation. The results using this method show a remarkable recognition rate of 96.0%.The proposed method provides a new way for vocabulary recognition.