Light Fixture Microphone Network for Speech Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data communication systems, particularly in free space optical communication, face challenges in distinguishing speech signals from ambient noise, especially when the noise includes speech-like interference, and require complex and power-intensive solutions for effective signal modulation and noise reduction.
Innovation Solution
A method utilizing a spatially distributed network of light transmitting apparatuses with microphones that employ beamforming and Green's Function processing to separate acoustic sources, incorporating propagation models that account for direct and indirect sound paths, and using neural networks or probabilistic approaches to enhance signal processing and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional speech enhancement techniques are used to reduce noise, then noise reduction is achieved, but the techniques are ineffective when noise consists of speech signals
Solution Approach 1:
The patent transitions from temporal/spectral signal processing to spatial processing by utilizing a distributed network of microphones in light transmitting apparatuses. By adding the spatial dimension through beamforming and Green's Function processing, the system can distinguish between desired speech sources and interfering speech sources based on their spatial locations, thereby resolving the limitation of conventional techniques that fail against speech-like noise.
Solution Approach 2:
The patent divides the acoustic environment into multiple spatial zones using a distributed microphone network. Each light transmitting apparatus acts as an independent sensing node, and through beamforming techniques, the system segments the sound field to identify and isolate specific speech sources from interfering speech sources based on their spatial separation.
2Measurement precision
If beamforming with multiple microphones is used to separate speech sources, then speech separation improves, but the system complexity and power consumption increase
Solution Approach 1:
The patent makes light transmitting apparatuses serve dual functions: providing illumination and capturing acoustic signals. By integrating microphones into existing lighting fixtures, the system achieves speech separation capabilities without adding dedicated microphone arrays, thereby reducing overall system complexity and leveraging existing infrastructure.
Solution Approach 2:
The system utilizes the existing spatial distribution of light transmitting apparatuses for acoustic sensing. Rather than requiring a separate dedicated microphone array installation, the lighting infrastructure itself provides the sensing function, reducing deployment complexity and utilizing already-positioned devices for dual purposes.
3Measurement precision
If fixed beamforming is used to enhance speech from a predetermined direction, then speech enhancement is achieved, but the listener must physically look at the target to obtain maximum amplification
Solution Approach 1:
The patent implements electronically-steerable beamforming that can dynamically adjust the beam direction without mechanical movement. The system can electronically steer the beam to track or follow speech sources, and can also provide omnidirectional coverage by processing signals from multiple apparatuses, allowing users to receive enhanced speech regardless of their physical orientation or line of sight to the speaker.
4Object-affected harmful factors
If adaptive beamforming is used to identify noise sources and steer nulls, then noise reduction improves, but the approach is not effective for wearable devices
Solution Approach 1:
The patent distributes microphones across multiple spatial locations in the environment rather than concentrating them in a single wearable device. This spatial distribution creates a large effective aperture for beamforming, enabling robust noise suppression and speech separation that would be difficult to achieve with the limited baseline available in wearable devices. The system trades local concentration for global distribution.
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 enables effective separation and communication of speech signals in complex environments with fewer microphones, reducing noise interference and maintaining low power consumption, while avoiding flicker and radio signal interference, and is adaptable for various settings including residential and commercial spaces.
Implementation Method 1
capturing, with a spatially distributed network of light transmitting apparatuses that include microphones, an acoustic audio input
Implementation Method 2
transmitting, using one or more of the light transmitting apparatuses, the processed audio input
Implementation Method 3
a body-worn receiver having at least one photo detector disposed on a surface of the body-worn receiver, the photo detector being operable to receive the modulated light intensity and convert the modulated light intensity into an electrical signal
Data Source
AI summary
A visible light audio system is operable to enable free space optical communication of audio signals via transmission of modulated light intensity at a light source to a photo diode being operably engaged with a demodulator and audio output device. Embodiments of the present disclosure enable a plurality of visible light transmitting apparatuses being installed in a commercial or residential dwelling and operably engaged over a network to combine their microphone inputs via spatial, amplitude, spectral, and/or temporal filtering and physical and geometrical modeling methods to separate one or more acoustic sources using a visible light audio system comprising an array of light sources being operable to receive an audio source input.