Voice conversation method, device, computer equipment and storage medium

Abstract

The present application relates to a voice conversation method, device, computer equipment and storage medium. The method includes: acquiring a target conversational text of speech to be generated; performing semantic recognition on the target conversational text, obtaining a semantic feature vector corresponding to a target word segment in the target conversation text, and determining based on the semantic feature vector corresponding to the target word segment The emotional prosody label corresponding to the target participle; determine the semantic boundary in the target conversational text, insert boundary text in the target conversational text based on the semantic boundary, obtain the first conversational text; use the first conversational text And the emotional prosody label corresponding to the target word segmentation is input into the speech synthesis model to synthesize the target conversational speech, wherein the speech synthesis model encodes the boundary text, and obtains the vector in the target conversational speech based on the encoded vector Voice pause information; output the target conversational voice. Adopting this method can improve the conversation effect.

Description

technical field [0001] The present application relates to the field of voice technology, in particular to a voice conversation method, device, computer equipment and storage medium. Background technique [0002] With the development of science and technology, speech synthesis technology has appeared. Speech technology can be used to synthesize speech based on text, making the conversation more efficient. For example, during a conversation, the text of the conversation can be obtained, the conversational voice is generated based on the text, and the conversational robot is controlled to emit the conversational voice, so that a voice conversation can be carried out through the conversational robot. However, the voice sent by the conversational robot is usually very blunt, resulting in Conversation is poor. Contents of the invention [0003] Based on this, it is necessary to provide a voice conversation method, device, computer equipment and storage medium for the above tec...

AI Technical Summary

Problems solved by technology

For example, during a conversation, the text of the conversation can be obtained, a conversational voice can be generated based on the text, and the conversational robot can be controlled to emit the conversational voice, so that the conversational robot can be used for a voice conversation. However, the voice of the conversational robot is usually very blunt, resulting in Poor conversation

Images