Smart Device Speech Control Without Wake-Up Words
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing smart device control methods require a wake-up word, leading to complex operations and poor user experience due to the need for multiple steps and potential power wastage.
Innovation Solution
A method that performs speech recognition on acquired speech signals to determine if the control instruction matches the current operation scene of the smart device, allowing for adjustments to the device's state without the need for wake-up words, thereby simplifying control and improving user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If wake-up word is required to control smart device, then device can be activated intentionally, but operation complexity increases and user experience deteriorates
Solution Approach 1:
The patent extracts the wake-up word requirement from the control process, allowing speech recognition to trigger device control directly without needing a separate wake-up word step. The speech recognition module continuously monitors speech signals and executes control instructions when confidence thresholds are met, eliminating the unnecessary wake-up word layer.
Solution Approach 2:
The system performs preliminary speech recognition and analysis before requiring user confirmation. The control instruction is pre-processed and validated against the current operation scene, so when the user confirms with a simple gesture or additional speech, the execution is immediate and accurate, reducing overall operation complexity.
2Speed
If wake-up state is maintained continuously, then device responds immediately to control instructions, but power consumption increases
Solution Approach 1:
The system uses periodic speech signal sampling and recognition instead of continuous wake-up state maintenance. The speech recognition module periodically checks for control instructions and only activates full processing when speech is detected, creating a rhythm of low-power monitoring followed by high-power processing only when needed.
Solution Approach 2:
The system dynamically adjusts its operational state based on detected speech signals. When no speech is detected, the system remains in low-power mode with minimal processing. When speech is detected and confidence threshold is met, the system transitions to active control mode, optimizing the balance between response speed and power consumption in real-time.
3Loss of time
If speech recognition is always active, then control instructions are detected immediately, but false activations and power waste occur
Solution Approach 1:
The system implements feedback mechanisms where speech recognition results are validated against multiple criteria including confidence thresholds, operation scene context, and user confirmation requirements. Only when all feedback checks pass is the control instruction executed, preventing false activations while maintaining responsive detection.
Solution Approach 2:
The system changes recognition parameters dynamically based on context. Confidence thresholds, sampling rates, and processing depth are adjusted according to the current operation scene and detected speech characteristics, enabling efficient detection with reduced power consumption by avoiding full processing of every speech signal.
4Device complexity
If control instruction matching is performed without scene awareness, then processing is simpler, but control accuracy and appropriateness decrease
Solution Approach 1:
The operation scene recognition module serves multiple functions: it contextualizes control instructions, validates their appropriateness, and adjusts system behavior accordingly. This single multi-functional component enables accurate scene-aware control without proportionally increasing overall system complexity.
Data Source
AI summary
Embodiments of the present disclosure provide a method for controlling a smart device, a computer device and a non-transitory computer readable storage medium. The method includes performing speech recognition on a speech signal acquired by the smart device; determining whether a control instruction corresponding to the speech signal matches with a present operation scene of the smart device; and adjusting an operation state of the smart device according to the control instruction when the control instruction matches with the present operation scene.