Two-Tier Keyword Detection for Noisy Audio Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice command systems face challenges in accurately detecting keywords due to variations in background noise, reverberation, and user accents, leading to high false acceptance and rejection rates.
Innovation Solution
A two-tiered system is employed, where a first detector device processes audio and determines a confidence value for keyword presence, and a second, more powerful detector device confirms the detection, allowing for refined keyword detection and adjustment of settings to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single detector device is used for keyword detection, then the device complexity is low, but the measurement precision of keyword detection is insufficient leading to high false acceptance and rejection rates
Solution Approach 1:
The keyword detection system is segmented into two distinct detector devices: a first detector device that performs initial keyword detection and generates a confidence value, and a second detector device that performs additional testing to confirm detection. This segmentation allows each detector to specialize in specific detection tasks, improving overall measurement precision while distributing system complexity across multiple components rather than concentrating all complexity in a single device.
2Reliability
If a second detector device with more processing power is added to confirm keyword detection, then the reliability of keyword detection is improved, but the device complexity increases
Solution Approach 1:
The first detector device acts as an intermediary between the audio input and the second detector device. It performs preliminary processing, generates a confidence value, and selectively forwards audio segments that meet confidence thresholds to the second detector device. This intermediary role reduces the processing burden on the second detector while maintaining high reliability through layered verification, thereby improving reliability without proportionally increasing overall system complexity.
3Speed
If the first detector device takes action immediately based on confidence value, then the response speed is fast, but the reliability of the action may be reduced without confirmation
Solution Approach 1:
The system dynamically adjusts its response strategy based on the confidence value generated by the first detector device. When the confidence value exceeds a predetermined threshold, the system takes immediate action without waiting for second detector confirmation, ensuring fast response speed. When the confidence value is below the threshold, the system waits for confirmation from the second detector device, ensuring action reliability. This dynamic adaptation allows the system to optimize both speed and reliability based on real-time detection confidence.
Data Source
AI summary
Systems, apparatuses, and methods are described for keyword detection. A first detector device may receive audio and process the audio to determine a confidence value regarding presence of a keyword. A copy of the audio may be passed to a second detector device to perform additional testing for the audio. The first detector device may take an action before or after sending to the second detector device based on the confidence value.