Voice Assistant State Management for Seamless Interaction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current voice interaction systems require users to constantly wake up a voice assistant using a wake-up word, leading to a non-smooth user experience and increased power consumption.
Innovation Solution
The system employs a low computing power voice recognition model to detect voice instructions without waking up the voice assistant, allowing for full-time wake-up-free voice interaction. When a matching voice instruction is detected, the system performs the corresponding operation and then wakes up the voice assistant for further interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the voice assistant is kept in wake-up state to enable continuous voice interaction, then the user experience is improved, but the power consumption increases
Solution Approach 1:
The system dynamically adjusts the voice assistant's operational state based on real-time voice detection needs. A lightweight voice detection model runs continuously in sleep state, and only transitions to full wake-up state when actual voice instructions are detected, optimizing the balance between interaction smoothness and power consumption
Solution Approach 2:
The voice recognition system is segmented into two distinct models: a lightweight detection model for continuous monitoring in sleep state, and a comprehensive recognition model for full voice interaction in wake-up state. This segmentation allows the system to use appropriate model complexity for each operational phase
2Measurement precision
If the voice assistant is woken up for every voice instruction, then the recognition accuracy is improved, but the user experience deteriorates due to frequent wake-up words
Solution Approach 1:
The lightweight voice detection model performs preliminary voice detection continuously in the background before full voice recognition is needed. This preliminary action identifies potential voice instructions ahead of time, allowing the system to wake up the voice assistant only when necessary, thereby maintaining recognition accuracy while improving user experience
3Use of energy by moving object
If a lightweight voice recognition model is used for continuous detection, then the power consumption is reduced, but the recognition capability is limited
Solution Approach 1:
The recognition system is divided into two specialized models: a lightweight detection model optimized for low-power continuous monitoring, and a comprehensive recognition model optimized for accurate voice instruction processing. Each model is tailored to its specific function, resolving the trade-off between power consumption and recognition capability
Solution Approach 2:
The lightweight detection model serves as an intermediary between continuous voice monitoring and full voice recognition. It filters and pre-processes voice inputs, passing only relevant detections to the comprehensive recognition model, thereby enabling the lightweight model to compensate for its limited recognition capability through intelligent intermediation
Data Source
AI summary
A voice interaction method and a related apparatus are provided. When a voice assistant is not woken up, an electronic device may recognize whether a detected voice matches a preset intention. If the voice matches the intention, the electronic device may perform an operation corresponding to the intention that matches the voice, and wake up the voice assistant. After waking up the voice assistant, the electronic device can more accurately respond to a subsequent request of a user. If there is no voice interaction in a preset time period after the voice assistant is woken up, the electronic device may switch the voice assistant from a wake-up state to a sleep state.