Method for detecting audio signal and apparatus
Embodiments disclosed herein provide a method for detecting an audio signal and an apparatus, where the method includes determining an input audio signal as a to-be-determined audio signal; determining an enhanced segmental signal-to-noise ratio (SSNR) of the audio signal, where the enhanced SSNR is greater than a reference SSNR; and comparing the enhanced SSNR with a voice activity detection (VAD) decision threshold to determine whether the audio signal is an active signal. According to the method and the apparatus provided in the embodiments, an active voice and an inactive voice can be accurately distinguished.
1. A method for detecting an audio signal, wherein the method is used by a signal detecting apparatus comprising a processor and a memory, and the method comprises:
determining that an input audio signal is an unvoiced signal;
determining an enhanced segmental signal-to-noise ratio (SSNR) of the audio signal, wherein the enhanced SSNR is greater than a reference SSNR; and
comparing the enhanced SSNR with a voice activity detection (VAD) decision threshold to determine whether the audio signal is an active signal,
wherein the determining the enhanced SSNR comprises:
determining the enhanced SSNR using the formula
SSNR′= x *SSNR+ y,
wherein SSNR indicates the reference SSNR, SSNR′ indicates the enhanced SSNR, and x and y indicate enhancement parameters.
2. The method according to claim 1 , wherein the input audio signal comprises 20 sub-bands, the 20 sub-bands are from sub-band 0 to sub-band 19, and sub-band SNRs of sub-band 18 and sub-band 19 are greater than a first preset threshold.
3. A method for detecting an audio signal, wherein the method is used by a signal detecting apparatus comprising a processor and a memory, and the method comprises:
determining that an input audio signal is an unvoiced signal;
determining a weight of each signal-to-noise ratio (SNR) of each sub-band in the audio signal, wherein a first weight of a SNR of a high-frequency end sub-band, wherein the SNR of the high-frequency end sub-band is greater than a first preset threshold, is greater than a second weight of a SNR of a second sub-band, wherein the second sub-band is one of sub-bands except the high-frequency end sub-band in the audio signal;
determining an enhanced segmental signal-to-noise ratio (SSNR) according to each SNR of each sub-band and each weight of each SNR of each sub-band in the audio signal, wherein the enhanced SSNR is greater than a reference SSNR; and
comparing the enhanced SSNR with a voice activity detection (VAD) decision threshold to determine whether the audio signal is an active signal.
4. The method according to claim 3 , wherein the input audio signal comprises 20 sub-bands, the 20 sub-bands are from sub-band 0 to sub-band 19, and sub-band SNRs of sub-band 18 and sub-band 19 are greater than the first preset threshold.
5. A signal detecting apparatus, wherein the apparatus comprises:
a memory storage comprising instructions; and
one or more processors in communication with the memory, wherein the one or more processors execute the instructions to:
determine that an input audio signal is an unvoiced signal;
determine an enhanced segmental signal-to-noise ratio (SSNR) of the audio signal, wherein the enhanced SSNR is greater than a reference SSNR; and
compare the enhanced SSNR with a voice activity detection (VAD) decision threshold to determine whether the audio signal is an active signal,
wherein the one or more processors execute the instructions to determine the enhanced SSNR using the formula
SSNR′= x *SSNR+ y,
wherein SSNR indicates the reference SSNR, SSNR′ indicates the enhanced SSNR, and x and y indicate enhancement parameters.
6. The apparatus according to claim 5 , wherein the input audio signal comprises 20 sub-bands, the 20 sub-bands are from sub-band 0 to sub-band 19, and sub-band SNRs of sub-band 18 and sub-band 19 are greater than a first preset threshold.
7. A signal detecting apparatus, wherein the apparatus comprises:
a memory storage comprising instructions; and
one or more processors in communication with the memory, wherein the one or more processors execute the instructions to:
determine that an input audio signal is an unvoiced signal;
determine a weight of each signal-to-noise ratio (SNR) of each sub-band in the audio signal, wherein a first weight of a SNR of a high-frequency end sub-band, wherein the sub-band SNR is greater than a first preset threshold, is greater than a second weight of a SNR of a second sub-band, wherein the second sub-band is one of sub-bands except the high-frequency end sub-band in the audio signal, and determine an enhanced segmental signal-to-noise ratio (SSNR) according to each SNR of each sub-band and each weight of each sub-band SNR of each sub-band in the audio signal, wherein the enhanced SSNR is greater than a reference SSNR; and
compare the enhanced SSNR with a voice activity detection (VAD) decision threshold to determine whether the audio signal is an active signal.
8. The apparatus according to claim 7 , wherein the input audio signal comprises 20 sub-bands, the 20 sub-bands are from sub-band 0 to sub-band 19, and sub-band SNRs of sub-band 18 and sub-band 19 are greater than the first preset threshold.