Voice print recognition method and device, storage medium and electronic equipment

Abstract

The invention provides a voice print recognition method and device, a storage medium and electronic equipment, and relates to the technical field of artificial intelligence. The method comprises the following steps: determining a candidate unit structure; constructing and training a neural network model; adding a mask module into the neural network model, and constructing a generator; using the trained neural network model as a baseline network, and constructing and training a generative adversarial network model through the baseline network, the generator and a discriminator; according to the trained generative adversarial network model, determining a to-be-clipped intermediate state node and clipping the to-be-clipped intermediate state node; after retraining the cut neural network model, extracting voice print features of the to-be-recognized target; and determining a voice print recognition result based on the similarity. The problems that computing resources are consumed in the searching process of an existing neural network architecture, the number of network parameters in the searched network structure is large, and the calculation amount is large can be solved.

Description

technical field [0001] The present disclosure relates to the technical field of artificial intelligence, and in particular, to a voiceprint recognition method, a voiceprint recognition device, a computer-readable medium, and electronic equipment. Background technique [0002] In recent years, Neural Architecture Search (NAS, Neural Architecture Search), as a technology that can automatically search for the best neural network architecture, has achieved great success in various tasks, such as image classification, semantic segmentation, object detection, etc. performance beyond that of artificially designed network architectures. [0003] The conventional NAS method selects the best candidate network architecture, but it consumes a lot of computing resources; the searched network units are simply spliced, and there are still a large number of network parameters in the entire network. [0004] In the existing network architecture search method, the search process needs to con...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the embodiments of the present disclosure is to provide a voiceprint recognition method, a voiceprint recognition device, a computer-readable medium, and an electronic device, thereby to a certain extent reducing the consumption of computing resources in the search process of the existing neural network architecture, and the searched In the network structure, the number of network parameters is large and the amount of calculation is large

Images