Audio Gain Control for Keyboard Noise During Speech Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automatic Gain Control (AGC) mechanisms in audio signals often mistakenly identify noise from input device activities, such as keyboard tapping, as speech, leading to inappropriate gain adjustments that can disrupt audio calls by either amplifying noise or failing to amplify speech sufficiently.
Innovation Solution
The solution involves analyzing input signals from devices like keyboards and mice to determine noise-generating activities, allowing the AGC to revert or delay gain adjustments made during such activities, thereby applying an unadjusted or modified gain that is closer to the previous speech-based gain, without complex signal processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the AGC mechanism uses broad characteristics to identify speech components, then more speech components are identified, but non-speech sounds (noise) are also mistakenly identified as speech
Solution Approach 1:
The patent segments the audio signal analysis into multiple independent characteristic checks (pitch frequency, spectral tilt, zero-crossing rate, signal strength) rather than relying on a single broad classification. This allows the system to evaluate each characteristic separately and make more accurate speech/non-speech determinations by considering the combination of features.
Solution Approach 2:
The patent introduces an intermediary noise detection mechanism that monitors for non-speech sounds separately from the speech identification process. This intermediary system can detect keyboard tapping and other noise patterns, then provide feedback to the AGC mechanism to prevent inappropriate gain adjustments based on noise alone.
2Reliability
If the AGC mechanism adjusts gain based on keyboard tapping noise, then the noise level is controlled, but speech amplification is disrupted
Solution Approach 1:
The patent implements a feedback mechanism where the AGC mechanism monitors the audio signal, identifies when keyboard tapping is occurring through noise detection, and adjusts gain accordingly. The system provides feedback to distinguish between noise-generated signals and speech, allowing it to maintain appropriate gain levels for speech while controlling noise amplification.
Solution Approach 2:
The patent applies preliminary anti-action by detecting keyboard tapping noise before it can cause problematic gain adjustments. The noise detection mechanism identifies noise patterns in advance and prevents the AGC from making inappropriate gain increases that would amplify the noise, thereby protecting the speech signal quality.
3Measurement precision
If the AGC mechanism increases gain to compensate for low speech signal, then speech becomes audible, but noise is also amplified
Solution Approach 1:
The patent applies dynamic gain adjustment where the gain level changes continuously based on the real-time characteristics of the audio signal. Rather than applying a fixed gain, the system dynamically adapts the gain level by evaluating speech characteristics (pitch frequency, spectral tilt, zero-crossing rate) and adjusting accordingly, allowing speech to be amplified when present while minimizing noise amplification when speech is absent or weak.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method and system for modifying an audio signal, the method comprising: receiving the audio signal at signal processing means; analysing the received audio signal to identify characteristic signal components in the audio signal; determining that another signal is input to the signal processing means, the input signal resulting from an activity which generates noise in the audio signal; and selectively applying an adjusted gain to the audio signal based on the determination that the input signal is input to the signal processing means, wherein the adjusted gain is generated in dependence upon the signal strength of the identified signal components.