A voice wake-up configuration method and system for a multi-model child intelligent seat device
By acquiring the seat model identification code, loading hardware parameters, and performing acoustic channel calibration and noise suppression, a model-adaptive wake-up word feature template is generated. This solves the consistency and anti-interference issues of voice wake-up systems for different seat models, achieving efficient voice wake-up recognition and low-power operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SUZHOU COOL BULL ELECTRONIC TECH CO LTD
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing voice wake-up systems for smart child car seats cannot effectively cope with the differences in hardware configurations between different seat models, resulting in poor wake-up recognition consistency and anti-interference capabilities, as well as a lack of adaptive capabilities, which affects user experience and battery life.
By acquiring the seat model identification code and loading hardware parameters, acoustic channel calibration and functional module noise suppression are performed to generate a model-adaptive wake-up word feature template. In low-power monitoring mode, frequency domain preprocessing and timing consistency verification are performed, and the wake-up threshold is dynamically adjusted to optimize wake-up recognition.
It has achieved adaptive capability of voice wake-up system across different seat models, improved recognition consistency and anti-interference ability, reduced R&D costs and power consumption, extended equipment uptime, and improved user experience.
Smart Images
Figure CN122417014A_ABST