Action Identification Device Noise Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for identifying user actions through sound face challenges in accurately distinguishing action sounds from mixed signals with noise, leading to decreased identification accuracy.
Innovation Solution
The proposed method involves acquiring sound data from a microphone, calculating feature amounts, determining user presence, calculating noise feature amounts when the user is not present, and subtracting these noise feature amounts from the sound data to extract action sound feature amounts when the user is present, thereby enhancing identification accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If noise reduction processing is applied to the voice noise mixed signal, then noise is reduced, but the action sound as an identification target is also reduced, making it difficult to accurately identify the action
Solution Approach 1:
The system performs preliminary noise measurement and characterization during periods when no user action is detected. The noise feature amount is calculated and stored in advance, then applied for noise removal when user actions occur. This preliminary action allows the system to have an accurate baseline of environmental noise without interfering with action sounds.
Solution Approach 2:
The system extracts only the noise component from the mixed signal by measuring it during periods when no action sound is present. By separating the noise measurement phase from the action detection phase, the system can extract pure noise features without contamination from action sounds, then subtract this extracted noise from the mixed signal during action detection.
2Object-affected harmful factors
If noise is reduced from the signal containing action sound and noise, then noise level decreases, but the action sound is also reduced, decreasing identification accuracy
Solution Approach 1:
The system performs preliminary noise characterization by measuring the noise feature amount during periods when no user is present or no action is occurring. This pre-measured noise profile is stored and then used as a reference for subtracting noise from signals containing action sounds, ensuring that noise reduction does not compromise action sound integrity.
Solution Approach 2:
The noise feature amount serves as an intermediary element that mediates between the noisy signal and the clean action sound. By introducing this intermediate noise characterization data, the system can selectively remove noise components while preserving action sound components, acting as a bridge that enables precise noise cancellation without affecting the target signal.
Data Source
AI summary
An action identification device acquires sound data from a microphone, calculates a feature amount of the sound data, determines whether or not a user is present in a space in which the microphone is installed, calculates a noise feature amount indicating a feature amount of noise based on the calculated feature amount and stores the calculated noise feature amount in a noise feature amount storage unit in a case where the user is not present in the space, subtracts the noise feature amount stored in the noise feature amount storage unit from the calculated feature amount to extract an action sound feature amount indicating a feature amount of an action sound generated by an action of the user in a case where the user is present in the space, and identifies an action of the user by using the action sound feature amount.