Multi-Speaker Voice Signal Analysis for Call Quality Metrics

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies lack comprehensive methods to analyze and evaluate the quality of telephone calls, including conversational dynamics, acoustic fidelity, and paralinguistic aspects, particularly in multi-channel and multi-speaker interactions, without being language-dependent.

Innovation Solution

A system and method for voice signal analysis on telephone calls that employs statistical methodologies and machine learning models to compute dialogue-related, acoustic quality, and paralinguistic metrics, using preprocessing, metrics computation, tagging, and alarming modules to assess and synthesize performance indicators, optionally generating reports.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If comprehensive quality attributes including conversational dynamics, acoustic fidelity, and paralinguistic aspects are analyzed, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvequality evaluation precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the voice signal analysis into three distinct computational dimensions: conversational dynamic analysis, acoustical fidelity assessment, and paralinguistic cues interpretation. Each dimension processes specific metrics independently, allowing comprehensive quality evaluation while managing complexity through modular organization of analysis tasks.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system transitions from traditional single-dimension voice quality assessment to a multi-dimensional evaluation framework that simultaneously analyzes conversational dynamics, acoustic properties, and paralinguistic features. This dimensional expansion enables comprehensive quality measurement without requiring a single overly complex analysis module.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If language-independent analysis is implemented, then adaptability is improved, but measurement precision may worsen

Engineering Contradiction:
Improvelanguage independenceVSAvoidanalysis accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system extracts and analyzes only those voice signal features that are universally present across all languages, such as acoustic properties, paralinguistic cues, and conversational dynamics. By separating language-specific content from language-independent structural and acoustic characteristics, the system achieves both language independence and maintained measurement precision.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system changes the analyzed parameters from language-specific textual content to language-independent acoustic and behavioral parameters. This parameter transformation allows the system to evaluate voice quality across different languages while maintaining consistent and precise measurement criteria based on universal voice signal characteristics.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If multiple metrics including dialogue-related, acoustic quality, and paralinguistic metrics are computed, then measurement precision is improved, but loss of information increases

Engineering Contradiction:
Improveevaluation comprehensivenessVSAvoiddata processing overhead
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The system merges the three distinct metric computation processes (conversational dynamics, acoustic quality, and paralinguistic cues) into a unified analysis framework that processes voice signals simultaneously across all dimensions. This consolidation reduces redundant data processing and information loss while maintaining comprehensive measurement precision through integrated metric computation.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP4641567A1System and method for voice signal analytic on telephone calls
Publication Date: 2025.10.29 BTS TECHNOLOGY SERVICES SA
  • EP4641567A1 patent drawingFigure 1~2
  • EP4641567A1 patent drawingFigure 3~5
  • EP4641567A1 patent drawingFigure 6

AI summary

A method for analyzing voice signals on telephone calls to assess the quality of telecommunication services comprehensively is disclosed. The method comprises the preprocessing stage (100), the metrics computation stage (200), the tagging stage (300). The metrics computation stage (200) is based on dialogue-related aspects of the conversation among speakers, the acoustic quality of the voice signal, and paralinguistic aspects of the speech segments within the voice signal. Optionally, the method comprises establishing alarms (400) and generating comprehensive reports (500). It is a multiple channel and speaker telephone conversation-oriented method that is language independent. This voice signal evaluation allows the online monitoring, including speech and non-speech segments during calls, as well as the offline retrospective evaluation of the telecommunication service segmented by criteria such as geographical regions or call center.