Voice interaction processing method and device, terminal and storage medium

Abstract

The invention discloses a voice interaction processing method and device, a terminal and a storage medium. The method comprises steps of obtaining a first voice interaction fragment; determining a voice recognition text corresponding to the first voice interaction fragment; performing semantic segment extraction processing on the speech recognition text through a semantic extraction model to obtain a first semantic segment; identifying the intention of the first semantic segment to obtain at least one intention identification result; and determining a to-be-executed target intention recognition result from the at least one intention recognition result. According to the method, adverse effects on natural language understanding recognition caused by the fact that excessive noise, namely useless input, enters a subsequent natural language understanding process are avoided, and accuracy of input voice intention recognition and execution in a streaming voice interaction scene is improved.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a voice interaction processing method, device, terminal and storage medium. Background technique [0002] The streaming dialogue method is one of the intelligent voice interaction methods. During the streaming dialogue process, the voice interaction system can continue to "speak" while continuing to monitor like a human without waking up with a wake-up word (that is, the voice interaction system completes the user's voice). The operation of feedback after input analysis), and continuous monitoring and "speaking" can overlap, which is more in line with the real human interaction form. [0003] In the streaming dialogue scenario, the user will continue to speak and input voice, and at the same time, the colloquial expression will contain many useless words, making the voice input noisy and prone to overly long input sentences. The lack of targeted processing leads to the...

AI Technical Summary

Problems solved by technology

[0003] In the streaming dialogue scenario, the user will continue to speak and input voice, and at the same time, the colloquial expression will contain many useless words, making the voice input noisy and prone to overly long input sentences. There is no targeted processing, resulting in the need for excessive processing of useless input in the subsequent natural language understanding process, which reduces the accuracy of input voice intent recognition and execution in streaming voice interaction scenarios

Images