Audio Noise Classification Using Temporal and Frequency Indicators

Solution Overview

Problem

Existing noise classification techniques for audio signals in telecommunications are complex and resource-intensive, requiring significant computing time and resources, making them unsuitable for real-time applications such as mobile communication.

Innovation Solution

A method that classifies background noise using only two indicators: a time indicator and a frequency indicator, calculated from the sound level variation and frequency spectrum of the noise signal, respectively, allowing for efficient classification with minimal computational burden and enabling real-time processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional noise classification methods using multiple indicators (e.g., 30 indicators with HMM) are employed, then classification accuracy is improved, but processing time and computational resources increase significantly

Engineering Contradiction:

Improveclassification accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts only the two most essential indicators (time indicator and frequency indicator) from the traditional 30 indicators used in HMM-based classification. This selective extraction maintains classification accuracy while dramatically reducing computational complexity and processing time, making the system suitable for real-time mobile communication applications

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces complex, resource-intensive classification models (HMM requiring 30 indicators) with a simpler, more efficient classification algorithm that uses only 2 indicators. This substitution uses cheaper computational resources while achieving comparable classification accuracy, enabling deployment in resource-constrained mobile devices

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

2Measurement precision

If complex noise classification techniques (e.g., PCA-based classification with 30 indicators) are used, then classification performance is improved, but device complexity and processing resources increase

Engineering Contradiction:

Improveclassification accuracyVSAvoidprocessing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the two most essential indicators (time indicator and frequency indicator) from the traditional 30 indicators used in HMM-based classification. This selective extraction maintains classification accuracy while dramatically reducing computational complexity and processing time, making the system suitable for real-time mobile communication applications

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of starting with many indicators and reducing them through complex dimensionality reduction techniques like PCA, the patent inverts the approach by directly selecting only the two most critical indicators from the beginning. This reverse engineering approach simplifies the classification system while maintaining effectiveness

Inventive Principle:

Principle #13The other way round (Inversion)

3Reliability

If existing noise classification methods are applied in real-time mobile communication, then communication quality can be maintained, but processing resources are excessively consumed

Engineering Contradiction:

Improvecommunication qualityVSAvoidprocessing energy

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent extracts only the two most essential indicators (time indicator and frequency indicator) from the traditional 30 indicators used in HMM-based classification. This selective extraction maintains classification accuracy while dramatically reducing computational complexity and processing time, making the system suitable for real-time mobile communication applications

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the parameters used for noise classification from 30 complex indicators to just 2 simplified indicators (time and frequency indicators). This parameter reduction significantly decreases computational energy consumption while maintaining the reliability needed for real-time mobile communication quality assurance

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS8972255B2Method and device for classifying background noise contained in an audio signal

Publication Date: 2015.03.03 ORANGE SA

AI summary

Embodiments of methods and devices for classifying background noise contained in an audio signal are disclosed. In one embodiment, the device includes a module for extracting from the audio signal a background noise signal, termed the noise signal. Also included is a second that calculates a first parameter, termed the temporal indicator. The temporal indicator relates to the temporal evolution of the noise signal. The second module also calculates a second parameter, termed the frequency indicator. The frequency indicator relates to the frequency spectrum of the noise signal. Finally, the device includes a third module that classifies the background noise by selecting, as a function of the calculated values of the temporal indicator and of the frequency indicator, a class of background noise from among a predefined set of classes of background noise.