Noise Reduction Apparatus Using Speech Segmentation and Direction Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise cancellation technologies are ineffective in high noise environments, failing to provide high-quality voice sound in wireless communication systems due to inadequate noise reduction capabilities.
Innovation Solution
A noise reduction apparatus comprising a speech segment determiner, a voice direction detector, and an adaptive filter that uses information from multiple microphones to identify speech segments and determine voice incoming direction, enabling effective noise reduction by subtracting noise components from voice signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If a noise signal obtained from a sub-microphone is subtracted from a voice signal obtained from a main microphone, then noise components are reduced, but the noise reduction function does not work well in high noise level environments
Solution Approach 1:
The patent segments the voice signal into speech segments and non-speech segments based on acoustic characteristics. The noise reduction processing is then applied selectively: strong noise reduction is applied to non-speech segments while preserving speech segments. This segmentation approach allows the system to maintain high noise reduction effectiveness in high noise environments by avoiding the removal of speech content during noise reduction processing.
Solution Approach 2:
The patent changes the parameter of noise reduction strength based on the detected speech segment status. When speech segments are detected, the noise reduction parameter is adjusted to preserve speech content while still reducing noise. This dynamic parameter adjustment enables effective noise reduction in high noise levels without degrading voice quality.
2Object-affected harmful factors
If noise reduction is applied to remove noise components, then voice quality improves, but speech segments may be erroneously removed in high noise environments
Solution Approach 1:
The patent segments the audio signal into speech segments and non-speech segments using acoustic feature analysis. By identifying which segments contain speech content and which contain only noise, the system can apply noise reduction selectively to non-speech segments only, preventing the loss of speech information while effectively removing noise components.
Solution Approach 2:
The patent uses feedback from acoustic feature analysis to control the noise reduction process. The speech segment detector continuously monitors the signal and provides feedback about the presence of speech segments, which then controls the noise reduction processing to avoid removing speech content while still eliminating noise.
3Device complexity
If conventional noise cancellation is used, then the system is simple, but it cannot adapt to high noise level environments
Solution Approach 1:
The patent introduces dynamic adaptability by continuously detecting acoustic features to identify speech segments and adjusting noise reduction processing accordingly. The system dynamically switches between different noise reduction strategies based on the detected environment, enabling adaptation to high noise levels while maintaining reasonable system complexity through software-based adaptation rather than hardware complexity.
Data Source
AI summary
It is determined whether or not a sound picked up by at least either a first microphone or a second microphone is a speech segment. When it is determined that the sound picked up by the first or the second microphone is the speech segment, a voice incoming direction indicating from which direction a voice sound travels is detected based on a first sound pick-up signal obtained based on a sound picked up by the first microphone and a second sound pick-up signal obtained based on a sound picked up by the second microphone. A noise reduction process is performed using the first and second sound pick-up signals based on speech segment information indicating that the sound picked up by the first or the second microphone is the speech segment and voice incoming-direction information indicating the voice incoming direction.