Voice output method, device, storage medium and electronic device
Through image acquisition and multimodal recognition technology, the voice assistant is automatically turned on and feedback on the user's action behavior is output, which solves the problem of wake-up words in the existing technology and improves the intelligence of the intelligent voice assistant.
CN111951787BActive Publication Date: 2025-03-18BEIJING XIAOMI PINECONE ELECTRONICS CO LTD
2 Cites 0 Cited by
Patent Information
- Application Number
- CN202010761629.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-07-31
- Publication Date
- 2025-03-18
- Estimated Expiration
- 2040-07-31
AI Technical Summary
Technical Problem
The existing voice assistant system requires the user to speak a fixed wake-up word before it can be turned on, which limits the way the user interacts with the voice assistant.
Method used
通过图像采集单元监测目标对象的动作行为,利用预先训练的多模态识别模型确定动作行为发生的目标场景和事件,从而自动开启语音助手系统,并根据场景和事件确定并输出相应的语音反馈。
Benefits of technology
It realizes that the user's action behavior can be responded to the user without wake-up words, improves the intelligence of the voice assistant and enhances the interaction ability with the user.
✦ Generated by Eureka AI based on patent content.
Abstract
The present disclosure is about a voice output method, device, storage medium, and electronic device. The method includes: monitoring the action behavior of the target object through an image acquisition unit to obtain image information containing the action behavior; determining the target scene where the action behavior occurs and the target event corresponding to the action behavior according to the pre-trained multimodal recognition model and the image information; after determining the target scene and the target event, controlling the voice assistant system to start; determining the target voice from the voice library of the voice assistant system according to the target scene and the target event; outputting the target voice, and the voice text corresponding to the target voice is the feedback content text for the action behavior. The voice assistant can be awakened in response to the action behavior of the target object, and the user's action behavior can be identified, and then the feedback voice for the action behavior can be output to actively interact with the user, thereby improving the intelligence of the intelligent voice assistant.
Need to check novelty before this filing date? Find Prior Art