Voice Detection Method Using Adaptive Thresholds in Noisy Environments

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing voice activity detection methods fail to accurately detect speech signals in noisy environments, such as sports events or construction sites, due to interference from background noise, leading to poor performance and false detections.

Innovation Solution

A voice detection method that involves sampling the acoustic signal into discrete frames, calculating a detection function based on a difference function, and adapting the threshold adaptively using reference values from sub-frames to minimize false detections and improve speech detection accuracy in noisy conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional voice activity detection methods are used in noisy environments, then the detection process is simple, but the detection accuracy deteriorates due to background noise interference

Engineering Contradiction:
Improvespeech signal detection accuracyVSAvoidbackground noise interference
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The acoustic signal is divided into multiple frames, and each frame is further subdivided into sub-frames. This segmentation allows the detection algorithm to analyze local characteristics within sub-frames and make frame-level decisions, improving robustness against noise by focusing on localized speech patterns rather than being affected by global noise conditions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The detection threshold is made adaptive rather than fixed. The threshold dynamically adjusts based on the estimated noise level in the current environment, allowing the system to maintain high detection accuracy across varying noise conditions. This is achieved through noise level estimation and threshold adaptation mechanisms that respond to changing acoustic environments.

Inventive Principle:
Principle #15Dynamics

2Reliability

If a fixed threshold is used for voice detection, then the detection algorithm is simple, but false detections increase in noisy environments

Engineering Contradiction:
Improvedetection reliabilityVSAvoidthreshold adaptation mechanism
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system implements feedback through adaptive threshold adjustment based on noise level estimation. The detected noise level feeds back into the threshold calculation, which then adjusts the detection threshold accordingly. This closed-loop approach improves reliability by continuously adapting to environmental conditions while maintaining manageable complexity through efficient estimation algorithms.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If noise reduction techniques are applied to improve speech detection, then detection accuracy improves, but processing time and computational load increase

Engineering Contradiction:
Improvespeech signal detection accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system applies partial noise reduction by focusing computational resources only on the necessary analysis - calculating detection functions for voice activity detection rather than performing full-spectrum noise reduction. This selective approach achieves sufficient detection accuracy without the excessive processing time required for comprehensive noise reduction techniques.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3078027B1Voice detection method
Publication Date: 2018.05.23 ADEUNIS RF
  • EP3078027B1 patent drawingFigure 1
  • EP3078027B1 patent drawingFigure 2~3
  • EP3078027B1 patent drawingFigure 4

AI summary

The invention relates to a voice detection method which makes it possible to detect the presence of voice signals in an noisy acoustic signal x(t) from a microphone, comprising the following consecutive steps: calculating a detection function FD(τ) based on calculating a difference function D(τ) varying in accordance with the shift τ on an integration window with length W starting at the time t0, with: a step of adapting the threshold in said current interval, in accordance with values calculated from the acoustic signal x(t) established in said current interval; searching for the minimum of the detection function FD(τ) and comparing said minimum with a threshold, for (τ) varying in a predetermined time interval referred to as current interval so as to detect the possible presence of a fundamental frequency F0 that is characteristic of a voice signal in said current interval.