Adaptive Audio Signal Analysis via Source Grouping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing techniques are inadequate for isolating sounds in complex audio environments with multiple sources and noise, leading to high word error rates and limited speech enhancement.

Innovation Solution

An adaptive multiple-model optimizer system that includes a segment grouping engine and a source grouping engine to generate source models from feature segments, allowing for the modification of audio signals by controlling noise and enhancing desired sounds.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If existing sound isolation techniques are used, then simple audio signals can be processed, but complex multi-source audio signals cannot be effectively isolated

Engineering Contradiction:

Improvecapability to handle complex audio environmentsVSAvoidsound isolation accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent segments the audio signal processing into distinct feature extraction, segment grouping, and source grouping stages. Feature segments are extracted from the audio signal, then grouped into segment models, and finally into source models. This segmentation allows the system to handle complex multi-source audio environments by processing different audio features separately and combining them systematically.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent employs dynamic model selection and adaptation where the system adjusts its processing based on the characteristics of the audio signal. The segment grouping engine and source grouping engine dynamically create and update models based on the extracted features, allowing the system to adapt to varying audio conditions and improve reliability in complex environments.

Inventive Principle:

Principle #15Dynamics

2Measurement precision

If simple feature extraction is used, then processing speed is maintained, but speech recognition accuracy deteriorates in noisy environments

Engineering Contradiction:

Improvespeech recognition accuracyVSAvoidprocessing system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments audio features into distinct categories (tone, transient, noise) and processes them through separate grouping engines. This segmentation of feature processing allows for more precise speech recognition by treating different audio characteristics differently, while the modular architecture manages the complexity through organized, separate processing stages.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces segment models as intermediary representations between raw feature extraction and final source identification. These segment models serve as intermediate structures that organize features before final source model creation, improving measurement precision by providing structured intermediate representations that capture speech characteristics more accurately.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Object-affected harmful factors

If traditional audio processing is used, then system simplicity is maintained, but noise reduction capability is insufficient

Engineering Contradiction:

Improvenoise reduction effectivenessVSAvoidaudio processing system complexity

Core Design Contradiction:

Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The patent extracts specific audio features (tone, transient, noise) from the mixed audio signal and processes them separately through the segment and source grouping engines. By taking out and separately processing noise features from the mixed signal, the system achieves effective noise reduction while managing complexity through focused, feature-specific processing stages.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system applies different processing qualities to different feature segments. Tone segments, transient segments, and noise segments are grouped and processed with appropriate models tailored to their specific characteristics. This local quality approach allows effective noise reduction by applying specialized processing to noise features while preserving speech features with appropriate processing.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS8315857B2Systems and methods for audio signal analysis and modification

Publication Date: 2012.11.20 SAMSUNG ELECTRONICS CO LTD

AI summary

Systems and methods for modification of an audio input signal are provided. In exemplary embodiments, an adaptive multiple-model optimizer is configured to generate at least one source model parameter for facilitating modification of an analyzed signal. The adaptive multiple-model optimizer comprises a segment grouping engine and a source grouping engine. The segment grouping engine is configured to group simultaneous feature segments to generate at least one segment model. The at least one segment model is used by the source grouping engine to generate at least one source model, which comprises the at least one source model parameter. Control signals for modification of the analyzed signal may then be generated based on the at least one source model parameter.