Table 3 Description of the top 12 vocal features
From: Differentiation between depression and bipolar disorder in child and adolescents by voice features
|  | Name of features | Meaning of features | − log (P) |
|---|---|---|---|
1 | pcm_RMSenergy_sma_rqmean | Represents the mean value of the root mean square (RMS) energy of a speech file | 314.97 |
2 | pcm_fftMag_spectralSlope_sma_amean | Represents the mean value of the slope of the power spectrum of the speech signal | 303.08 |
3 | audSpec_Rfilt_sma[1]_percentile1.0 | Represents the energy magnitude of the signal in the 301–600 Hz frequency band in hundredths of a decimal place | 303.01 |
4 | pcm_fftMag_spectralHarmonicity_sma_amean | Represents the average value of the spectral harmonicity of the audio signal | 297.07 |
5 | pcm_fftMag_spectralFlux_sma_percentile1.0 | Represents the 1.0 percentile of the spectral flux of the audio signal | 294.77 |
6 | pcm_fftMag_fband250-650_sma_percentile1.0 | Represents the 1.0 percentile of the spectral amplitude of the audio signal in the frequency band 250Â Hz to 650Â Hz | 294.17 |
7 | pcm_fftMag_fband1000-4000_sma_percentile1.0 | Represents the 1.0 percentile of the spectral amplitude of the audio signal in the frequency band 1000Â Hz to 4000Â Hz | 289.21 |
8 | audspec_lengthL1norm_sma_amean | Represents the average of the L1 parametres of the spectral length of the audio signal | 288.56 |
9 | pcm_RMSenergy_sma_percentile1.0 | Represents the 1.0 percentile of the root-mean-square energy of the audio signal | 288.3 |
10 | pcm_fftMag_fband250-650_sma_de_quartile2 | Represents the dichotomous difference of the spectral amplitude of the audio signal in the frequency band 250Â Hz to 650Â Hz | 280.2 |
11 | pcm_fftMag_fband250-650_sma_amean | Represents the average value of the spectral amplitude of the audio signal in the frequency band 250Â Hz to 650Â Hz | 279.83 |
12 | audspec_lengthL1norm_sma_percentile1.0 | Represents the 1.0 percentile of the L1 parity of the spectral length of the audio signal | 279.23 |