Audio Nuisance Notification via Probabilistic Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In audio communication scenarios, unwanted sounds such as breath sounds, keyboard typing, and finger tapping are difficult to mitigate without compromising voice quality, as conventional noise suppression techniques are ineffective for rapidly varying nuisances, leading to uncomfortable communication experiences for other users.
Innovation Solution
A method and system that determine the probability of nuisance presence in an audio signal using features like spectral difference, signal-to-noise ratio, spectral centroid, and power difference, and track the signal over multiple frames to notify the user of their unwanted sounds, allowing them to adjust and minimize the noise without degrading voice quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-generated harmful factors
If conventional noise suppression techniques are used, then constant noises can be removed, but rapidly varying nuisances cannot be effectively removed
Solution Approach 1:
The system dynamically adapts its noise suppression approach by detecting whether the nuisance is constant or rapidly varying through statistical analysis of audio features over time. The notification mechanism is activated dynamically based on the duration and characteristics of the detected nuisance, allowing the system to adjust its behavior to match the specific type of interference encountered.
Solution Approach 2:
The system changes its operational parameters by switching between different detection and suppression strategies based on the temporal characteristics of the nuisance. When rapid variation is detected, the system transitions to a notification-based approach rather than traditional suppression, effectively changing the control parameter from active suppression to user awareness prompting.
2Object-generated harmful factors
If noise suppression techniques are applied to remove nuisances, then unwanted sounds can be reduced, but voice quality may be compromised
Solution Approach 1:
The system implements feedback by notifying the user about the detected nuisance, allowing them to self-correct their behavior. This feedback loop avoids the need for aggressive noise suppression that would degrade voice quality, as the user becomes aware of and can voluntarily stop the nuisance-causing behavior such as keyboard typing or finger tapping.
Solution Approach 2:
The notification system enables self-service by empowering the user to identify and eliminate their own nuisances. Instead of the system forcibly suppressing sounds (which would affect voice quality), the user themselves takes action to stop the unwanted behavior after being notified, thus preserving voice quality while removing nuisances.
3Ease of operation
If notifications are presented to users about their nuisances, then user experience can be improved, but false notifications may cause unnecessary interruptions
Solution Approach 1:
The system performs preliminary analysis by tracking audio features over multiple frames and comparing them against established thresholds before issuing a notification. This preliminary action ensures that only genuine nuisances that meet specific criteria trigger notifications, reducing false positives while maintaining user awareness of actual problems.
Solution Approach 2:
The system replaces simple threshold-based detection with a more sophisticated statistical analysis approach, substituting mechanical decision-making with probabilistic evaluation. By analyzing the distribution and temporal characteristics of audio features, the system can more reliably distinguish between actual nuisances and normal variations in speech or environment, thereby improving notification accuracy.
Data Source
AI summary
Example embodiments disclosed herein relate to audio signal processing. A method of indicating a presence of a nuisance in an audio signal is disclosed. The method includes determining a probability of the presence of the nuisance in a frame of the audio signal based on a feature of the audio signal, the nuisance representing an unwanted sound made by a user, in response to the probability of the presence of the nuisance exceeding a threshold, tracking the audio signal based on a metric over a plurality of frames following the frame, determining, based on the tracking, that the presence of the nuisance is to be indicated to the user, and in response to the determination, presenting to the user a notification of the presence of the nuisance. Corresponding system and computer program product are also disclosed.