Audio Feature Extraction for ACR Database Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The increasing number of content pieces handled by Automatic Content Recognition (ACR) services using only audio information leads to a significant increase in database capacity and operational burdens, including increased processing time and maintenance costs, as the data amount of feature point information grows.

Innovation Solution

An information processing device and method that extracts feature point information only from the main sound of audio content, reducing the need for extensive database capacity and operational resources by focusing on the main sound, even when sub sounds are listened to, thus minimizing database preparation and maintenance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If feature point information is extracted from all audio channels including sub sounds, then content identification accuracy is improved, but database capacity and operational burdens increase significantly

Engineering Contradiction:
Improvecontent identification accuracyVSAvoiddatabase capacity
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts only the necessary component (main sound feature point information) from the complete audio content, excluding sub sounds and other non-essential audio channels. This selective extraction reduces the quantity of data stored in the database while maintaining sufficient accuracy for content identification, directly resolving the contradiction between identification reliability and database capacity.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments audio content into distinct components (main sound and sub sounds) and processes only the main sound for feature point extraction. By dividing the audio content and selecting only the essential segment for identification purposes, the system reduces database burden while preserving identification accuracy.

Inventive Principle:
Principle #1Segmentation

2Reliability

If feature point information is extracted from all audio channels including sub sounds, then content identification accuracy is improved, but processing time and operational costs increase

Engineering Contradiction:
Improvecontent identification accuracyVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts only the essential main sound component for feature point extraction, eliminating the need to process sub sounds and other redundant audio channels. This extraction approach significantly reduces processing time while maintaining identification accuracy, as the main sound contains sufficient information for reliable content recognition.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies partial action by processing only the necessary portion (main sound) of the complete audio content rather than all audio channels. This partial processing approach is sufficient for achieving accurate content identification while dramatically reducing the time and computational resources required.

Inventive Principle:
Principle #16Partial or excessive action

3Quantity of substance

If feature point information is extracted only from main sound, then database capacity is reduced, but content identification may fail when sub sounds are listened to

Engineering Contradiction:
Improvedatabase capacityVSAvoidaudio channel compatibility
Core Design Contradiction:
Quantity of substanceVSAdaptability or versatility

Solution Approach 1:

The patent introduces a switching mechanism that acts as an intermediary between the user's audio channel selection and the feature point extraction process. When sub sounds are selected for listening, the switch automatically extracts feature points from the main sound instead, ensuring identification functionality is maintained across different audio channel selections while keeping the database compact.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent creates a universal identification system that works regardless of which audio channel the user is listening to. By designing the system to always reference main sound feature points even when sub sounds are played, the system achieves multi-functional adaptability - supporting different listening preferences while maintaining a single, reduced-capacity database.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP2782336B1Information processing device and information processing method
Publication Date: 2020.10.07 SATURN LICENSING LLC
  • EP2782336B1 patent drawingFigure 1
  • EP2782336B1 patent drawingFigure 2
  • EP2782336B1 patent drawingFigure 3

AI summary

An ACR service is realized by an information provision server which has not prepared feature point information of a sub sound because only the feature point information is extracted from a main sound and an inquiry is made to the information provision server even when the sub sound is listened to in a client device. In addition, even when content that has a main sound and a plurality of pieces of audio information is distributed, it is not necessary for the information provision server to prepare feature point information of sub sounds, and the capacity of the database may not increase.