Voice conversion method and device, equipment and storage medium

Abstract

The invention provides a voice conversion method and apparatus, a device and a storage medium. The method comprises the steps of determining a first low-dimensional vector corresponding to a target voice based on the target voice; separating the first low-dimensional vector to obtain a target identity voice vector; and determining a final conversion voice based on the sample identity voice vector and the target identity voice vector. According to the method, one-to-many voice conversion can be realized without adding an additional conversion model for the system, and meanwhile, a mode of obtaining the target voice by calculating the Mel-frequency cepstrum coefficient MFCC feature offset is adopted, so that a large amount of parallel corpora are not needed as training data, and the difficulty of collecting the training data is reduced; the method can be applied to the customer service platform, the one-to-many voice conversion technology is adopted to increase the types of convertible voices of the customer service system aiming at the problem that the voice conversion function of the customer service platform is single in voice, and the purpose of protecting the privacy of customers is achieved; the system is convenient and simple, has powerful functions, and can effectively and quickly realize voice conversion.

Description

technical field [0001] The invention relates to the field of voice conversion, in particular to a voice conversion method, device, equipment and storage medium. Background technique [0002] In recent years, artificial intelligence technology has developed rapidly, and the development of artificial intelligence has also brought many conveniences to people's daily life. Speech conversion, as a new technology in artificial intelligence, has a wide range of applications in the fields of TV movie dubbing, artificial intelligence (AI) anchor, privacy protection and other fields. Although voice conversion technology appeared for a short time, the development speed of voice conversion is very fast. After three technical iterations, it has entered the golden age of development. At present, speech conversion technology is close to natural human voice in terms of fluency and naturalness. [0003] In the customer service platform, since all voice calls are public, the effect of prote...

AI Technical Summary

Problems solved by technology

[0005] The current voice conversion function based on the customer service system is based on one-to-one voice conversion, and the voice conversion is relatively monotonous. If you want to add voice conversion, you need to add an additional model, which will take up more space in the system, and each model must be independent. training, training takes a lot of time

At present, there is no technical solution that can solve the above technical problems. Specifically, there is no voice conversion method, device, equipment and storage medium.

Images