Acoustic Signal Processing for Double-Talk Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing acoustic communication systems face challenges in reliably detecting double-talk, which complicates the removal of echo signals and degrades acoustic signal transmission.

Innovation Solution

An acoustic signal processing system comprising a frequency band reservation module and a double-talk indicator module, connected with an adaptive filter and a non-linear processor, which suppresses specific frequency bands to differentiate between echo and near-end speech, enabling effective double-talk detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-generated harmful factors

If traditional echo cancellation techniques are used, then echo removal is achieved, but double-talk detection becomes difficult and unreliable

Engineering Contradiction:
ImproveechoVSAvoiddouble-talk detection accuracy
Core Design Contradiction:
Object-generated harmful factorsVSMeasurement precision

Solution Approach 1:

The patent segments the acoustic signal into multiple frequency bands using filter banks. By analyzing energy distribution across different frequency bands separately, the system can distinguish between echo (which typically occupies specific frequency ranges) and near-end speech (which has different spectral characteristics), thereby enabling reliable double-talk detection while maintaining echo cancellation performance.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing strategies to different frequency bands. By examining the energy distribution and temporal characteristics in specific frequency regions, the system can locally identify double-talk conditions without being affected by echo in other bands, thus improving detection accuracy while preserving overall echo removal capability.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If frequency band reservation is applied, then double-talk detection accuracy is improved, but system complexity increases

Engineering Contradiction:
Improvedouble-talk detection accuracyVSAvoidsignal processing structure
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses filter banks to divide the signal into multiple frequency bands, which can be implemented using efficient digital signal processing algorithms. This segmentation approach improves detection accuracy by enabling frequency-selective analysis while maintaining computational feasibility through standard DSP techniques.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts energy information from specific frequency bands and uses this extracted feature for double-talk detection. By focusing only on the relevant energy distribution characteristics in selected frequency regions rather than processing the entire spectrum, the system achieves high detection accuracy with reduced computational complexity.

Inventive Principle:
Principle #2Taking out (Extraction)

3Object-generated harmful factors

If adaptive filtering is used for echo cancellation, then echo removal effectiveness is improved, but detection of near-end speech during double-talk becomes unreliable

Engineering Contradiction:
ImproveechoVSAvoidnear-end speech detection
Core Design Contradiction:
Object-generated harmful factorsVSReliability

Solution Approach 1:

The patent combines adaptive filtering for echo cancellation with frequency-band-based detection. The adaptive filter operates to remove echo from the composite signal, while simultaneously, the system analyzes energy distribution in specific frequency bands to detect near-end speech. This segmented approach allows reliable near-end speech detection during double-talk even in the presence of adaptive filtering operations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces frequency band energy analysis as an intermediary mechanism between the adaptive filter output and the double-talk detection decision. By examining the energy distribution in specific frequency bands as an intermediate step, the system can reliably detect near-end speech components that might otherwise be obscured by the adaptive filtering process.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP2779610B1Acoustic signal processing system capable of detecting double-talk and method
Publication Date: 2022.11.16 SEMICON COMPONENTS IND LLC
  • EP2779610B1 patent drawingFigure 1
  • EP2779610B1 patent drawingFigure 2~3
  • EP2779610B1 patent drawingFigure 4

AI summary

An acoustic signal processing system and method. In accordance with an embodiment, the acoustic signal processing system includes an adaptive filter that filters a signal from a frequency band reservation module and generates a filter signal that is received by a subtractor. The subtractor generates an error signal that is used by a double-talk indicator module to generate a control signal that indicates the presence of double-talk.