Audio data generating method and system for voice synthesis

Abstract

The invention provides an audio data generating method for voice synthesis. The audio data generating method comprises the following steps: extracting the text features in the text data to obtain textfeature data; carrying out the acceleration conversion processing on the text feature data through a neural network structure, and converting the text feature data into acoustic feature data; performing sound synthesis or selected splicing according to the acoustic feature data to obtain audio data. According to the invention, through the special anti-convolution structure is adopted, a good voice synthesis effect can be achieved on the premise that no any auto-regressive structure is included and few parameters are used, the calculation delay can be reduced while the prediction precision ofthe acoustic features can be guaranteed through the neural network structure, and the requirement for computing resources is reduced, the concurrent quantity is increased, the voice synthesis speed isincreased, and contribution is made to improvement of human-computer interaction experience.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to a method and system for generating audio data for speech synthesis. Background technique [0002] For a voice-based real-time human-computer interaction system, in order to achieve the optimal human-computer interaction experience, the time from the end of the user's voice pronunciation to the start of the machine's voice reply is called "response time". In order to achieve the optimal human-computer interaction experience, the total time of these three steps should be around 600ms. For most voice human-computer interaction systems, the information processing process needs to go through three steps of ASR-NLP-TTS in sequence. However, at present, the delay of a large number of high-naturalness TTS interfaces is above 150ms, which severely compresses the processing time of the other two steps (ASR, NLP), and thus limits the complexity and accuracy of information processing ...

AI Technical Summary

Problems solved by technology

However, at present, the delay of a large number of high-naturalness TTS interfaces is above 150ms, which severely compresses the processing time of the other two steps (ASR, NLP), and thus limits the complexity and accuracy of information processing in the other two steps. To improve the human-computer interaction experience, it is necessary to increase the speed of TTS or speech synthesis

Images