Segmented Spectrogram Code Detection for Audio Signal Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for detecting segments with specific sound signals in longer stored sound signals are inefficient, especially when dealing with background music, as they require extensive calculations and are slow due to the assumption of similar segments being identical to the reference signal, failing to account for overlapping noises and lacking speed in detection.

Innovation Solution

A detection system that encodes and compares small-region spectrograms of both the reference and stored signals using similarity-based codes, skipping unnecessary calculations by indexing and table referencing to quickly identify similar segments, reducing the need for histogram overlapping ratio calculations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If the time-series active search method is used to detect similar segments, then the detection speed is improved, but the accuracy deteriorates when overlapping noises are present

Engineering Contradiction:
Improvedetection speedVSAvoiddetection accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The patent divides the stored sound signal into multiple small segments and compares each segment with the reference signal. This segmentation allows the system to efficiently search through large databases by breaking down the complex comparison task into smaller, manageable units, thereby maintaining high detection speed while improving accuracy even in the presence of overlapping noises.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If the self-optimized spectral correlation method is used to detect background music, then the detection accuracy is improved, but the detection time increases significantly

Engineering Contradiction:
Improvedetection accuracyVSAvoiddetection time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by using simplified comparison criteria for initial screening of segments. Instead of performing exhaustive spectral correlation analysis on all possible segments, the system first identifies candidate segments using faster comparison methods, then applies more rigorous analysis only to promising candidates. This approach maintains high detection accuracy while dramatically reducing overall detection time.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If exhaustive comparison of all segments is performed to ensure high detection accuracy, then the detection precision is improved, but the computational complexity and time increase

Engineering Contradiction:
Improvedetection precisionVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary actions by pre-processing the sound signals to extract key features and create compact representations before the actual comparison process. This preliminary feature extraction reduces the dimensionality of the data and highlights the most discriminative characteristics, enabling accurate segment detection with significantly reduced computational complexity during the main detection phase.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS7860714B2Detection system for segment including specific sound signal, method and program for the same
Publication Date: 2010.12.28 NIPPON TELEGRAPH & TELEPHONE CORP
  • US7860714B2 patent drawing
  • US7860714B2 patent drawing
  • US7860714B2 patent drawing

AI summary

The present invention is a detection system of a segment including specific sound signal which detects a segment in a stored sound signal similar to a reference sound signal, including: a reference signal spectrogram division portion which divides a reference signal spectrogram into spectrograms of small-regions; a small-region reference signal spectrogram coding portion which encodes the small-region reference signal spectrogram to a reference signal small-region code; a small-region stored signal spectrogram coding portion which encodes a small-region stored signal spectrogram to a stored signal small-region code; a similar small-region spectrogram detection portion which detects a small-region spectrogram similar to the small-region reference signal spectrograms based on a degree of similarity of a code; and a degree of segment similarity calculation portion which uses a degree of small-region similarity and calculates a degree of similarity between the segment of the stored signal and the reference signal.