Cotalker Nulling via Multi Super Directional Beamformer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition systems face challenges in accurately distinguishing and suppressing co-talker speech, which can confuse the recognition process and reduce accuracy.
Innovation Solution
The use of spatially-selective microphone detection beams and directional lobes, achieved through digitally delaying signals from multiple microphones to create constructive and destructive interference patterns, allows for the selective capture of speech from a driver while suppressing co-talker speech by combining signals in a way that cancels out passenger audio.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional speech recognition systems are used, then speech recognition can be performed, but co-talker speech cannot be effectively suppressed leading to reduced recognition accuracy
Solution Approach 1:
The system segments the audio space into multiple directional beams using a microphone array. Each beam is directed toward a specific spatial location, allowing the system to separately process and identify speech signals from different directions. This segmentation enables the system to distinguish between the target speaker and co-talker by spatial separation.
Solution Approach 2:
The system creates directionally-selective beams with different spatial characteristics. Each beam has a specific directional pattern optimized for capturing speech from a particular direction while rejecting speech from other directions. This local quality differentiation in spatial selectivity allows the system to preferentially capture the target speaker's voice while suppressing co-talker interference.
2Object-affected harmful factors
If multiple microphones are used to create directional beams, then co-talker suppression capability is improved, but device complexity increases
Solution Approach 1:
The system replaces complex mechanical acoustic filtering structures with electronic signal processing. Instead of using physical barriers or complex microphone positioning mechanisms, the invention uses digital signal processing techniques to create directional beams and suppress co-talker speech. This substitution of electronic processing for mechanical solutions reduces overall system complexity while maintaining effective co-talker suppression.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively locates and suppresses co-talker speech, enhancing the accuracy of speech recognition systems by isolating the driver's speech and reducing interference from passenger audio, thereby improving overall recognition accuracy.
Implementation Method 1
achieved through digitally delaying signals from multiple microphones to create constructive and destructive interference patterns
Data Source
AI summary
Speech from a driver and speech from a passenger in a vehicle is selected directionally using a plurality of directional microphones. Sounds detected as coming from a passenger from a plurality of directional microphones are suppressed from sounds detected as coming from a driver by a second plurality of directional microphones.