Audio Scene Change Detection Using Feature Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting scene change points in audio signals, especially in transition scenes with gradual changes, are inefficient, leading to difficulties in accurately identifying desired scenes and increasing user burden in navigating through multiple potential change points.
Innovation Solution
An audio processing device that calculates feature data for unit sections of an audio signal, determines similarity sections, and prioritizes boundaries as scene change points based on boundary information, allowing for effective detection of scene change points even in transition scenes with gradual audio changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single specific method is used to detect scene change points, then the detection process is simplified, but the ability to detect a wide variety of different scene change points deteriorates
Solution Approach 1:
The audio signal is divided into multiple unit sections, and scene change detection is performed separately for each section using similarity comparison with reference sections. This segmentation allows the system to handle diverse scene change types through multiple localized detection operations rather than a single complex method.
Solution Approach 2:
The system changes detection parameters dynamically by adjusting reference section selection and similarity thresholds based on the characteristics of each unit section. This allows adaptation to different scene change types (sudden transitions, gradual transitions, etc.) while maintaining a unified detection framework.
2Reliability
If all boundaries are detected as scene change points, then comprehensive detection is achieved, but user burden increases due to multiple potential change points
Solution Approach 1:
The system applies different priority levels to different boundaries based on local characteristics such as similarity degree and boundary strength. High-priority boundaries (clear scene changes) are distinguished from low-priority boundaries (subtle transitions), allowing users to focus on significant scene changes while reducing navigation burden.
Solution Approach 2:
The similarity section concept acts as an intermediary between raw boundary detection and final scene change identification. By introducing this intermediate layer that evaluates boundary quality and priority, the system filters out insignificant boundaries while maintaining comprehensive detection of meaningful scene changes.
3Ease of manufacture
If conventional threshold-based detection is used, then the method is simple to implement, but scene change points in transition scenes with gradual changes are difficult to detect
Solution Approach 1:
The system performs preliminary similarity analysis between unit sections and reference sections before making scene change decisions. This preliminary action of comparing audio characteristics and establishing similarity relationships enables accurate detection of gradual scene changes while maintaining a simple threshold-based decision framework.
Solution Approach 2:
The reference section selection and similarity evaluation are dynamically adjusted based on the local audio characteristics of each unit section. This dynamic adaptation allows the system to maintain high detection accuracy for gradual scene changes while preserving the simplicity of the overall detection methodology.
Data Source
AI summary
An audio processing device including a feature calculation unit, a boundary calculation unit and a judgment unit, detects points of change of audio features from an audio signal in an AV content. The feature calculation unit calculates, for each unit section of the audio signal, section feature data expressing features of the audio signal in the unit section. The boundary calculation unit calculates, for each target unit section among the unit sections of the audio signal, a piece of boundary information relating to at least one boundary of a similarity section. The similarity section consists of consecutive unit sections, inclusive of the target unit section, which each have similar section feature data. The judgment unit calculates a priority of each boundary indicated by one or more of the pieces of boundary information and judges whether the boundary is a scene change point based on the priority.