Speech synthesis method and device based on low-resource language, equipment and storage medium

Abstract

The invention relates to an artificial intelligence technology, and discloses a speech synthesis method based on a low-resource language, which comprises the following steps: determining a low-resource language and a high-resource language, and obtaining a text corresponding to the low-resource language to obtain a low-resource language text; converting the low-resource language text into a low-resource language phoneme text; translating the low-resource language phoneme text into a high-resource language phoneme text by using a translation model trained based on dual learning; and performing speech synthesis on the high-resource language phoneme text by using a pre-trained speech synthesis model to obtain language speech. In addition, the invention also relates to a block chain technology, and the low-resource language text can be stored in a node of a block chain. The invention further provides a speech synthesis device based on the low-resource language, electronic equipment and a computer readable storage medium. The invention can provide a speech synthesis method for improving the speech synthesis effect and aiming at low-resource languages.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to a speech synthesis method, device, electronic equipment and computer-readable storage medium based on a low-resource language. Background technique [0002] Text-to-speech (TTS) has developed rapidly in recent years and has attracted extensive attention from academia and industry. The demand for speech synthesis in all walks of life is getting higher and higher. For example, the customer service industry can use speech synthesis to complete self-service voice services. [0003] There are more than 6,000 languages ​​in the world, and traditional speech synthesis methods cover only dozens of languages. The reason behind this cannot be separated from the difficulty of data set development. The development of a new language data set usually requires the hiring of professional voice actors for the language. The language customizes a large amount of high-qualit...

AI Technical Summary

Problems solved by technology

[0003] There are more than 6,000 languages ​​in the world, and traditional speech synthesis methods cover only dozens of languages. The reason behind this cannot be separated from the difficulty of data set development. The development of a new language data set usually requires the hiring of professional voice actors for the language. The language customizes a large amount of high-quality voice data, and then uses these high-quality voice data as a training data set to train the model. However, for many low-resource languages, such as dialects and other minority languages, there is usually no international phonetic alphabet or even the phonetic alphabet of the language. At the same time, the process of hiring linguistic experts to customize and analyze technical data such as the pronunciation of the language and phonemes and tones needs to consume a lot of money and time. Therefore, the training data set is very small

However, the effect of directly using the traditional speech synthesis model for speech synthesis of low-resource languages ​​is poor. Currently, there is no speech synthesis method specifically for low-resource languages.

Images