Voice Noise Reduction Using Dynamic Energy Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing noise reduction methods often result in unnatural voice output when low voltage signals are incorrectly identified as noise, necessitating an improvement in amplitude adjustment techniques to differentiate between noise and normal voice.

Innovation Solution

A method that divides input voice into segments, calculates a reference ratio based on maximum energy values and a predetermined energy value, and adjusts energy using a constraint coefficient to distinguish between noise and normal voice, ensuring natural voice output.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If low voltage signals are directly insulated as noise signals, then noise reduction effectiveness is improved, but voice naturalness deteriorates due to incorrect identification of normal voice as noise

Engineering Contradiction:

Improvenoise reduction effectivenessVSAvoidvoice naturalness

Core Design Contradiction:

Object-affected harmful factorsVSReliability

Solution Approach 1:

The patent changes the parameter used for noise identification from simple amplitude threshold to energy-based metrics. By calculating energy values of voice segments and comparing them against dynamically adjusted reference values, the system can distinguish between low-amplitude noise and low-amplitude normal voice, resolving the contradiction between noise reduction effectiveness and voice naturalness

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent introduces dynamic reference values that adapt to changing voice characteristics. The reference energy value is updated based on previous voice segments, allowing the system to dynamically adjust what constitutes 'noise' versus 'normal voice'. This dynamic adaptation prevents incorrect identification and maintains natural voice quality while reducing actual noise

Inventive Principle:

Principle #15Dynamics

2Object-affected harmful factors

If amplitude adjustment is applied to all low voltage signals, then noise interference is reduced, but voice quality deteriorates due to over-adjustment of normal voice segments

Engineering Contradiction:

Improvenoise interferenceVSAvoidvoice quality

Core Design Contradiction:

Object-affected harmful factorsVSManufacturing precision

Solution Approach 1:

The patent applies local quality by treating different voice segments differently based on their energy characteristics. Instead of uniform amplitude adjustment, the system calculates energy for each segment and applies adjustment only when the segment's energy is significantly below the dynamic reference value, preserving the quality of normal voice segments while reducing noise

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent implements feedback through continuous monitoring of voice segment energy and comparison against dynamically updated reference values. The system uses previous voice segments to adjust future reference values, creating a feedback loop that prevents over-adjustment and maintains voice quality by adapting to the actual voice characteristics

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS9514765B2Method for reducing noise and computer program thereof and electronic device

Publication Date: 2016.12.06 AIROHA TECHNOLOGY CORPORATION

AI summary

A method for reducing noise is used to divide a received voice into plural voice segments and set a predetermined energy value. The energy of voice segment which is higher than the predetermined energy value is determined as normal voice and outputs directly, and the energy of voice segment which is lower than the predetermined energy value is determined as noise and will be processed.