Adaptive Filter Pitch Extraction Under Background Noise

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio processing systems face challenges in accurately estimating the pitch of speech signals, especially in noisy environments where background noises obscure the signal, leading to decreased intelligibility and poor perceptual quality.

Innovation Solution

A system and method for extracting pitch from speech signals using adaptive filtering techniques, including a delay unit, adaptive filter, and weighting logic, which estimate the pitch by approximating the position of peaks in adaptive filter coefficients and improve perceptual quality by enhancing selected voiced and unvoiced segments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If adaptive filtering is used to extract pitch from speech signals, then pitch estimation accuracy is improved, but system complexity increases

Engineering Contradiction:
Improvepitch estimation accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The speech signal is divided into multiple frames, with each frame processed independently through the adaptive filtering pipeline (delay unit, adaptive filter, weighting logic, peak detection). This segmentation allows the complex pitch extraction task to be broken down into manageable sequential operations, improving accuracy through frame-by-frame analysis while keeping each processing stage relatively simple.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The adaptive filter coefficients are dynamically adjusted for each frame based on the speech signal characteristics. The system adapts the filtering parameters (delay samples, filter order, weighting factors) to match the varying pitch and spectral properties of different speech segments, thereby improving pitch estimation accuracy across diverse speech conditions without requiring a fixed complex structure.

Inventive Principle:
Principle #15Dynamics

2Reliability

If background noise is reduced to improve speech intelligibility, then speech clarity is improved, but signal processing complexity increases

Engineering Contradiction:
Improvespeech intelligibilityVSAvoidsignal processing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The adaptive filter acts as an intermediary between the noisy speech signal and the pitch extraction process. By introducing this intermediate filtering stage with adjustable coefficients and weighting logic, the system selectively enhances pitch-related frequency components while attenuating background noise, thereby improving speech intelligibility without requiring direct manipulation of the entire signal spectrum.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system changes key parameters (filter coefficients, weighting factors, delay samples) adaptively based on the input signal characteristics. This dynamic parameter adjustment allows the processor to optimize noise reduction and pitch extraction for different speech and noise conditions, improving intelligibility while avoiding the need for fixed complex processing structures.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP1796078B1Adaptive filter pitch extraction
Publication Date: 2009.01.07 QNX SOFTWARE SYSTEMS WAVEMAKERS INC
  • EP1796078B1 patent drawingFigure 1
  • EP1796078B1 patent drawingFigure 2~4
  • EP1796078B1 patent drawingFigure 5

AI summary

An enhancement system extracts pitch from a processed speech signal. The system estimates the pitch of voiced speech by deriving filter coeffcients of an adaptive filter and using the obtained filter coefficients to derive pitch. The pitch estimation may be enhanced by using various techniques to condition the input speech signal, such as spectral modification of the background noise and the speech signal, and/or reduction of the tonal noise from the speech signal.