In-Vehicle Speech Pickup Using Virtual Microphone Beamforming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing in-vehicle communication systems face challenges in maintaining a high Signal-to-Noise (SN) ratio for speech voices, leading to difficulties in clear audio transmission between seats, especially when the audio device output mixes with microphone outputs, causing reverberation and reduced audibility.
Innovation Solution
The system employs a signal processing unit that utilizes virtual microphones and adaptive filters to enhance the directivity of speech voice collection and output, improving the SN ratio without the need for additional microphones at the first seat. This is achieved by converting the speech voice collected by one microphone into the voice that would be collected by a virtual microphone closer to the speech position, and then combining the outputs of the actual and virtual microphones to form a high-directivity virtual microphone array.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a high-directivity microphone is used to improve the SN ratio of speech voice, then the SN ratio improves, but the device complexity and design limitation increase due to the large and unique shape
Solution Approach 1:
The patent creates a virtual microphone that copies the acoustic characteristics of a high-directivity microphone positioned at the first seat, using signal processing to simulate its output. This allows the system to achieve high-directivity performance without physically installing a large, complex microphone at the constrained location.
Solution Approach 2:
The patent introduces a virtual microphone as an intermediary element that mediates between the available microphones at the second seat and the desired high-directivity performance. The virtual microphone is synthesized through signal processing to provide the acoustic field information that would be obtained from a physical high-directivity microphone at the first seat.
2Measurement precision
If a microphone array is used to implement high-directivity microphone, then the directivity improves, but the cost and processing load increase due to the increase in the number of microphones
Solution Approach 1:
Instead of deploying multiple physical microphones to create a microphone array, the patent copies the directional response characteristics of a high-directivity microphone through signal processing. The virtual microphone synthesizes the directional pattern that would result from a physical array, achieving the same acoustic field estimation without the hardware overhead.
Solution Approach 2:
The patent replaces the mechanical system of multiple physical microphones with an electronic signal processing system. Rather than using additional hardware elements to achieve directionality, the system uses computational methods to synthesize the directional response, substituting electronic processing for mechanical expansion.
3Adaptability or versatility
If the audio device output is collected and mixed with the microphone output, then the communication coverage improves, but the SN ratio of speech voice decreases and reverberation increases
Solution Approach 1:
The patent extracts the speech voice signal from the mixed audio environment by using the virtual microphone to selectively capture and estimate the directional speech component. The signal processing isolates the speech signal from the reverberant and mixed audio content, separating the useful speech information from the harmful background noise and reverberation.
Solution Approach 2:
The patent converts the harmful effect of audio mixing and reverberation into a benefit by using the virtual microphone's directional characteristics to identify and extract the speech signal even in the presence of these adverse conditions. The system leverages the spatial information to distinguish speech from other audio sources, turning the complex mixed environment into an opportunity for selective signal extraction.
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 configuration effectively increases the directivity of the speech voice output towards the speech position, improving the SN ratio and ensuring clearer audio transmission to other seats within the vehicle, while maintaining a relatively efficient processing load.
Implementation Method 1
a filter that converts the first seat voice collected by the second microphone into the first seat voice collected by a first seat virtual microphone that is a virtual microphone located at a position closer to the first seat than the second microphone
Implementation Method 2
generates an output of a virtual microphone array including the first microphone and the first seat virtual microphone and having higher directivity to the speech position of the user in the first seat as compared with the first microphone
Data Source
Figure 1
Figure 2B~2A2
Figure 3~4
AI summary
Provided is an "in-vehicle communication support system" that improves an SN ratio of a speech voice output from a speaker with a relatively efficient configuration. A right seat processing unit (11) includes, assuming that a voice of a user in a right front seat is a right front seat voice, a filter HR (111) that converts the right front seat voice being output from a left front seat microphone (PL) disposed on a headrest of a left front seat into the right front seat voice collected by a right seat virtual microphone (PVR) that is a virtual microphone located on a left side of a headrest of the right front seat, a delay unit Z-TR (112) that delays and outputs an output from a right front seat microphone (PR) located on a right side of the headrest of the right front seat, a filter WRA (113) that extracts the right front seat voice being output from the filter HR (111), a filter WRB (114) that extracts the right front seat voice being output from the delay unit Z-TR (112), and a right adder (115) that adds outputs of the filter WRA (113) and the filter WRB (114) and outputs an added output as a right front seat speech voice signal (SR).