Configurable Voice Capture Front-End for Headsets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice capture algorithms for headsets are designed for specific headset form factors and microphone configurations, making it difficult to manage and tune voice capture solutions for a wide range of headset geometries, leading to performance degradation and increased costs due to the need for bespoke algorithms for each configuration.
Innovation Solution
A configurable signal processing device with a microphone signal router, voice activity detection modules, and a spatial noise reduction module using a generalized sidelobe canceller, which can adapt to different headset form factors and microphone positions through training and configuration, allowing for flexible routing and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voice capture algorithms are designed specifically for each headset form factor and microphone configuration, then voice capture performance is optimized for that specific configuration, but the complexity of managing and tuning multiple bespoke algorithms increases significantly
Solution Approach 1:
The patent applies universality by creating a single voice capture algorithm that can function across multiple headset form factors and microphone configurations. The system uses a configuration database that stores geometric parameters for different headset types, allowing the same core algorithm to adapt its behavior based on the detected headset configuration, thereby eliminating the need for separate bespoke algorithms for each form factor.
Solution Approach 2:
The patent implements parameter changes by dynamically adjusting algorithm parameters based on the headset configuration. The system detects the headset form factor and microphone positions, then modifies processing parameters such as beamforming weights, noise reduction thresholds, and signal routing configurations to optimize performance for the specific geometry without requiring a completely different algorithm for each case.
2Measurement precision
If bespoke voice capture algorithms are developed for each headset configuration, then optimal performance is achieved for that configuration, but engineering time and development costs increase
Solution Approach 1:
The patent applies preliminary action by pre-configuring a database of headset geometric parameters and algorithm settings for various known headset form factors. During product development, the system can quickly match a new headset configuration to the database and retrieve appropriate parameters, eliminating the need for time-consuming bespoke algorithm development and extensive tuning for each new headset design.
Solution Approach 2:
The system uses copying by replicating proven algorithm configurations from the database for similar headset types. When a new headset configuration is introduced, the system can copy parameters from a database entry that matches the form factor and microphone arrangement, providing a working baseline configuration immediately rather than starting from scratch.
3Measurement precision
If voice capture algorithms are optimized for specific headset geometries, then sound quality and noise reduction are maximized for that geometry, but the algorithms perform markedly degraded when the headset form factor differs from the expected geometry
Solution Approach 1:
The patent implements dynamics by making the algorithm configuration dynamic rather than static. The system detects the actual headset form factor and microphone positions at runtime, then dynamically adjusts processing parameters including beamforming directions, noise reduction aggressiveness, and signal routing to match the detected geometry, allowing optimal performance across varying headset configurations.
Solution Approach 2:
The patent applies local quality by tailoring specific processing parameters to local geometric characteristics of the headset. Different sections of the algorithm are adjusted based on local factors such as microphone spacing, earcup geometry, and head-to-headset distance, allowing the system to optimize noise reduction and voice capture for the specific physical arrangement without compromising overall adaptability.
Data Source
AI summary
A signal processing device for configurable voice activity detection. A plurality of inputs receive respective microphone signals. A microphone signal router configurably routes the microphone signals. At least one voice activity detection module receives a pair of microphone signals from the router, and produces a respective output indicating whether speech or noise has been detected by the voice activity detection module in the respective pair of microphone signals. A voice activity decision module receives the output of the voice activity detection module(s) and determines whether voice activity exists in the microphone signals. A spatial noise reduction module receives microphone signals from the microphone signal router, and performs adaptive beamforming based in part upon the output of the voice activity decision module, and outputs a spatial noise reduced output. The device permits simple configurability to deliver spatial noise reduction for one of a wide variety of headset form factors.