A digital human instruction interaction word recognition system and method based on FunASR

The FunASR-based digital human command interaction system solves the problems of insufficient adaptability of wake-up mechanisms and lack of semantic understanding capabilities, and realizes flexible wake-up word recognition and accurate transcription of interactive command words, thereby improving the user experience.

CN122417028APending Publication Date: 2026-07-17JIANGSU ZHUODUN INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-05-07
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing digital human command interaction technologies suffer from insufficient adaptability of wake-up mechanisms, limited interaction sentence structures, and a lack of semantic understanding capabilities, resulting in low accuracy in wake-up word recognition and transcription of interaction command words, and a poor user experience.

Method used

The system employs a FunASR-based digital human command interaction system. It generates standard speech signals through signal acquisition and preprocessing modules, combines a custom wake-up word library and a multi-level command recognition library, and utilizes the FunASR model for wake-up word recognition and command word transcription. It includes support for error correction and escaping algorithms and a large intelligent dialogue model, achieving flexible adaptation and accurate recognition.

Benefits of technology

It improves the flexibility and accuracy of wake word recognition, enhances the accuracy and response efficiency of interactive command words, reduces the cost of interactive memory, and improves the user experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122417028A_ABST
    Figure CN122417028A_ABST
Patent Text Reader

Abstract

本发明提供一种基于FunASR的数字人指令交互词识别系统及方法,涉及指令交互词识别技术领域,其系统包括信号获取与预处理模块,用于获取用户语音信号并进行预处理生成标准语音信号;唤醒词扫描与识别模块,用于调用自定义唤醒词库,基于FunASR模型扫描并识别标准语音信号对应的自定义唤醒词库中的唤醒词;指令词转写与数字人触发模块,用于唤醒词识别成功后,系统自动切换至指令生成模式,并调用多级指令识别库,基于FunASR模型对标准语音信号进行指令词转写,生成交互指令词并触发数字人交互机制,从而可以对唤醒词和交互指令词进行适配识别与精准转写,提高数字人交互的准确性、灵活性与响应效率,满足用户使用体验感。
Need to check novelty before this filing date? Find Prior Art