A new patent for honorable voice print has been announced, which can accurately identify the voice sent by wearing a mask.
CTOnews.com January 3 news, Glory Terminal Co., Ltd. today announced a new patent for "voiceprint recognition method and electronic equipment", which can accurately recognize the voice sent by wearing a mask.
According to the patent specification, when the user wears a mask, the mask will suppress the user's voice to a certain extent, and the user's wake-up voice is quite different from the registered voice when wearing the mask. As a result, electronic devices may not be able to recognize the user's own voice, and the voiceprint recall rate will be reduced, affecting the user's experience.
In order to solve the above technical problems, the application provides a voiceprint recognition method and electronic equipment to improve the voiceprint recall rate and enhance the user experience.
CTOnews.com learned that the patent summary shows that the method includes:
Feature extraction of the target speech, generation of the first target spectrogram, and feature extraction of the registered speech to generate the second target spectrogram
The target mask is generated according to the first target spectrogram, and the target mask is applied to the second target spectrogram to obtain the third target spectrogram.
The first voiceprint vector is extracted from the first target spectrogram and the second voiceprint vector is extracted from the third target spectrogram.
If the first and second voiceprint vector similarity is greater than or equal to the preset similarity threshold, it is determined that the target voice belongs to the user corresponding to the registered voice.
In the patent application, the masking effect of the mask on speech is equivalent to a mask acting on the voice, and the mask can flexibly and accurately simulate the inhibitory effect of the mask on sound at different frequencies, so that the voice emitted by the user wearing a mask can be more accurately identified and the recall rate of voiceprint can be improved.