Training method and device of voiceprint recognition model, electronic equipment and storage medium

Abstract

The invention discloses a training method and device of a voiceprint recognition model, electronic equipment and a storage medium. The method incudes: acquiring a training set which comprises user audio data; inputting the acoustic feature vector corresponding to each audio data into a voiceprint recognition model to obtain the acoustic feature vector corresponding to each audio data in the training set and predicted probability distribution; calculating cross entropy loss according to the predicted probability distribution corresponding to each audio data in the training set and the real probability distribution corresponding to each audio data; calculating central loss according to the voiceprint feature vector corresponding to each audio data in the training set and the reference voiceprint feature vector corresponding to user identification; updating the model parameters of the voiceprint recognition model according to the cross entropy loss and the central loss. By the training method, the voiceprint feature distance of the same user can be lowered while the voiceprint feature distance of different users can be increased.

Description

technical field [0001] The invention relates to the technical field of voiceprint recognition, in particular to a training method, device, electronic equipment and storage medium for a voiceprint recognition model. Background technique [0002] The training method of the traditional voiceprint recognition model generally adopts the training method of cross-entropy loss. The training process is as follows: first extract the acoustic features of the audio data used for training, input the acoustic features into the deep learning network, and obtain which user the audio data belongs to. , and then update the model parameters according to the error backpropagation between the predicted result and the real result (that is, the pre-labeled user on the audio data). [0003] The training method of cross-entropy loss only takes into account one goal of voiceprint recognition, which is to increase the distance between voiceprint features of different users as much as possible, but the...

AI Technical Summary

Problems solved by technology

[0004] Embodiments of the present invention provide a training method, device, electronic equipment, and storage medium for a voiceprint recognition model to solve the problem that the cross-entropy loss training method in the prior art ignores the goal of reducing the distance between the acoustic features of the same user, and it is difficult to improve The problem of the recognition accuracy of the voiceprint recognition model

Images