Approbation of the LS method for voice morphing

The article examines the method of least squares (hereinafter MLS) in the future in order to be able to use it to change and form the emotions of a human voice (robot). The optimal values of the dimensions of the order of the polynomial and the number of samples in each sound recording window have been determined by the matching criterion of the specified and received voice amplitudes of the voice. The results can be listened to and compared with the original voices. On the basis of the created base of voices of emotions, the coefficients of MLS of the sample have been calculated. The reverse calculation for these coefficients has been checked. Further, using the first non-emotional record, emotions are restored by coefficients. By the same coefficients, emotions are restored for another speaker of the same emotion phrase. The evaluation has been carried out on the basis of the method developed by the authors, as the sum of divisions for the given and obtained samples of emotions.