Adaptive Voice Detection Using Dual Integrators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Voice-detection devices face challenges in distinguishing human speech from background noise in noisy environments, making it difficult to activate microphones only in response to human voice and deactivate in its absence.
Innovation Solution
A voice detection system utilizing a first integrator with a fast attack time and a second integrator with a slower attack time, along with a comparator, to differentiate between human voice and ambient noise by analyzing energy changes across spectral bands, ensuring the microphone is activated only when a human voice is detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a threshold is set to activate voice-detection devices, then the microphone can be activated in response to human voice, but it becomes difficult to distinguish between human voice and background noise in noisy environments
Solution Approach 1:
The patent segments the voice detection process into multiple independent integrators, each with different attack times. The first integrator has a fast attack time for rapid response to voice, while the second integrator has a slow attack time to filter out background noise. This segmentation allows the system to distinguish between voice and noise by comparing the outputs of the two integrators.
Solution Approach 2:
The patent changes the attack time parameter of the integrators to achieve different response characteristics. By setting the first integrator with a fast attack time and the second with a slow attack time, the system creates a parameter difference that enables discrimination between voice signals and background noise, improving detection accuracy in noisy environments.
2Speed
If a fast attack time is used for voice detection, then the microphone responds quickly to human voice, but it may activate in response to transient background noise
Solution Approach 1:
The system segments the detection function into two integrators with different attack times. The fast-attacking first integrator ensures quick response to voice, while the slow-attacking second integrator acts as a filter that only activates after sustained voice input, preventing false activations from transient noise.
Solution Approach 2:
The second integrator with slow attack time serves as an intermediary element between the fast-attacking first integrator and the final detection decision. It mediates by requiring sustained energy levels before triggering activation, thus filtering out false positives while maintaining fast response to actual voice.
3Reliability
If a slow attack time is used for voice detection, then background noise is filtered out, but the microphone response to human voice is delayed
Solution Approach 1:
The patent divides the detection system into two integrators with different attack times. The first integrator with fast attack time provides rapid initial response to voice, while the second integrator with slow attack time provides noise filtering. Both integrators work in parallel, so the fast response is not compromised while maintaining noise filtering capability.
Solution Approach 2:
The system merges the fast response capability of the first integrator with the noise filtering capability of the second integrator by combining their outputs through a comparator. This combination allows the system to achieve both fast response and reliable noise filtering simultaneously.
Data Source
AI summary
A system for detecting a voice signal includes: a first integrator for receiving an input signal and for providing a first integrator output signal, wherein the first integrator includes a first attack time; a second integrator for receiving the input signal and for providing a second integrator output signal, the second integrator including a second attack time that is substantially slower than the first attack time; and a comparator configured for receiving the first and second integrator output signals and for providing a comparator output signal indicating detection of a voice signal when the first integrator output signal exceeds the second integrator output signal by at least a threshold amount.