Far-Field Signal Separation With Dynamic Algorithm Switching

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing far-field voice processing systems face challenges in effectively separating intended voice signals from background noise in varying environments, leading to degraded performance in speech-based activities due to the computational intensity of using both beamforming and blind source separation algorithms simultaneously.

Innovation Solution

A dynamic algorithm selection mechanism that transitions between beamforming and blind source separation algorithms based on environmental factors, reducing computational resources by selecting the optimal algorithm for the current noise conditions, and smoothing transitions to minimize audio artifacts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If both beamforming and blind source separation algorithms are used simultaneously to separate voice signals from background noise, then speech-based activity performance is improved, but computational overhead increases significantly

Engineering Contradiction:

Improvespeech-based activity performanceVSAvoidcomputational overhead

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system dynamically transitions between beamforming and blind source separation algorithms based on real-time environmental noise conditions. The signal separation module monitors noise levels and automatically switches algorithms to optimize performance while minimizing computational resource usage, rather than running both algorithms simultaneously or using a fixed algorithm.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the algorithm selection parameter based on environmental noise conditions. When noise levels exceed a threshold, the system transitions from beamforming to blind source separation, and vice versa. This parameter-based dynamic selection resolves the contradiction by adapting computational resource usage to actual performance needs.

Inventive Principle:

Principle #35Parameter changes

2Use of energy by moving object

If a single signal separation algorithm is used to reduce computational overhead, then computational resources are optimized, but performance degrades in varying noise environments

Engineering Contradiction:

Improvecomputational resourcesVSAvoidperformance in varying environments

Core Design Contradiction:

Use of energy by moving objectVSAdaptability or versatility

Solution Approach 1:

The system employs dynamic algorithm selection that adapts to changing environmental conditions. The signal separation module continuously monitors noise levels and transitions between beamforming and blind source separation algorithms accordingly, ensuring optimal performance across varying noise environments while maintaining efficient computational resource usage.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system uses feedback from environmental noise monitoring to dynamically adjust algorithm selection. The signal separation module receives feedback about current noise conditions and automatically selects the appropriate algorithm, creating a closed-loop system that maintains both computational efficiency and adaptability to varying environments.

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If algorithm transitions are implemented dynamically based on noise conditions, then adaptability to environmental changes is improved, but audio artifacts may be introduced during transitions

Engineering Contradiction:

Improveadaptability to noise conditionsVSAvoidaudio artifacts

Core Design Contradiction:

Adaptability or versatilityVSObject-generated harmful factors

Solution Approach 1:

The system applies smoothing transitions when switching between beamforming and blind source separation algorithms. This cushioning approach gradually blends the transition between algorithms rather than abrupt switching, minimizing the introduction of audio artifacts while maintaining adaptability to changing noise conditions.

Inventive Principle:

Principle #11Beforehand cushioning (Prior cushioning)

Solution Approach 2:

The system uses a smoothing mechanism as an intermediary during algorithm transitions. This intermediary process gradually mixes the outputs of both algorithms during the transition period, reducing abrupt changes and minimizing audio artifacts while still achieving adaptability to environmental noise conditions.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12603097B2Dynamic selection of appropriate far-field signal separation algorithms

Publication Date: 2026.04.14 AVAGO TECHNOLOGIES INTERNATIONAL SALES PTE LTD

AI summary

An exemplary implementation includes a non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform operations including receiving a first audio data, processing the first audio data by a first signal separation algorithm, in response to an output of the first signal separation algorithm satisfying at least one parameter, outputting the processed first audio data. In response to the output of the first signal separation algorithm not satisfying the at least one parameter selecting a second signal separation algorithm, which is different than the first signal separation algorithm, receiving a second audio data subsequent in time to receiving the first audio data, processing the second audio data by the second signal separation algorithm, and outputting the processed second audio data.