Adaptive Speech Filter for Ambient Noise Attenuation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for attenuating ambient noise in video recordings, such as spectral subtraction and beam forming, often require user interaction, are limited by computing capacity, and primarily operate in either the frequency or time domain, failing to effectively differentiate noise from speech without distorting the signal.

Innovation Solution

An adaptive speech filter system that analyzes input signals using recursive spectral analysis to identify energy-rich spectral components, determine their harmonic affiliation, and classify noise versus speech through zero-point variance analysis, applying a transfer function in the frequency domain with pre- and post-ringing filters to minimize distortion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If spectral subtraction or beam forming methods are used to attenuate ambient noise, then noise reduction capability is improved, but user interaction requirements increase or device complexity increases

Engineering Contradiction:
Improveambient noise attenuationVSAvoiduser interaction requirement
Core Design Contradiction:
Object-affected harmful factorsVSEase of operation

Solution Approach 1:

The system automatically identifies and processes noise segments without requiring user interaction. The computer-implemented method autonomously analyzes the audio signal, detects noise portions, generates filter curves, and applies noise attenuation, making the system self-sufficient and eliminating the need for manual noise sample selection or parameter adjustment.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The invention introduces an automatic noise detection and filtering system that acts as an intermediary between the raw audio signal and the final output. This intermediary system includes components for automatic noise segment identification, filter curve generation, and signal processing, which together automate the noise attenuation process without requiring direct user intervention.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Object-affected harmful factors

If multiple simultaneously recorded input signals from differently positioned microphones are used for active noise suppression, then noise attenuation effectiveness is improved, but device complexity and computational requirements increase

Engineering Contradiction:
Improveambient noise attenuationVSAvoidmicrophone array requirement
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The invention extracts and isolates the noise component from the audio signal by automatically identifying noise segments through spectral analysis. By separating the noise portion from the speech portion and processing them differently, the system achieves effective noise attenuation using a single microphone input, eliminating the need for complex multi-microphone arrays.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system dynamically adjusts filtering parameters based on the detected noise characteristics. By analyzing the spectral properties of noise segments and generating adaptive filter curves, the system changes filtering parameters in real-time to match the specific noise conditions, achieving effective attenuation without requiring multiple physical microphones.

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If automatic noise detection and removal is implemented, then ease of operation is improved, but computing capacity requirements and memory requirements increase

Engineering Contradiction:
Improveautomatic processing capabilityVSAvoidcomputing capacity requirement
Core Design Contradiction:
Ease of operationVSUse of energy by moving object

Solution Approach 1:

The audio signal is divided into discrete segments or frames for analysis. By processing the signal in smaller time-frequency segments rather than as a whole, the system reduces the computational burden on each processing step while maintaining effective noise detection and attenuation across the entire audio signal.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system applies noise attenuation selectively only to identified noise segments rather than processing the entire audio signal uniformly. By detecting and targeting only the portions containing noise while preserving speech segments, the system reduces overall computational requirements compared to applying heavy processing to the complete signal.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3032536B1Adaptive speech filter for attenuation of ambient noise
Publication Date: 2020.04.01 BELLEVUE INVESTMENTS GMBH & CO
  • EP3032536B1 patent drawingFigure 1
  • EP3032536B1 patent drawingFigure 2
  • EP3032536B1 patent drawingFigure 3

AI summary

According to a preferred aspect of the instant invention, there is provided a system and method that allows the user to attenuate ambient noise in speech recordings in the audio part of a video recording. The user does not need to define particular sections or samples or individual parameters. The system is automatically analyzing the input signal and in a plurality of individual steps detects the ambient noise, determines an adaptive filter, implements the filter and therewith attenuates the ambient noise accordingly.