Video-Assisted Audio Object Extraction for Immersive Channel-Based Audio

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Traditional channel-based audio content formats lack the capability to provide immersive experiences similar to object-based audio, necessitating the extraction of audio objects from channel-based content to align with video content for enhanced immersion.

Innovation Solution

A method and system for video content-assisted audio object extraction, where video object-based information is used to assist in extracting audio objects from channel-based audio content, utilizing techniques like video object extraction, audio template generation, and metadata estimation to improve the precision and alignment of audio object extraction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If audio objects are extracted from channel-based audio content without video assistance, then the extraction process is simpler, but the precision and alignment of audio object extraction deteriorates

Engineering Contradiction:
Improveaudio object extraction precisionVSAvoidextraction system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses video content as an intermediary to assist audio object extraction. Video objects and their metadata (position, velocity, size) serve as mediating information that guides the extraction of corresponding audio objects from channel-based audio, improving extraction precision without requiring complex audio-only processing

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent transitions from one-dimensional audio signal processing to two-dimensional video-audio joint processing. By incorporating video spatial information (x, y coordinates) and temporal information, the system achieves more precise audio object localization and extraction than audio-only methods

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If traditional channel-based audio formats are used, then compatibility with existing systems is maintained, but the immersive listening experience deteriorates

Engineering Contradiction:
Improveimmersive experience capabilityVSAvoidaudio format complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the audio signal into distinct audio objects with associated metadata (position, velocity, size, type), separating them from the background audio bed. This segmentation enables independent control and spatial positioning of individual sound sources, creating immersive 3D audio experiences while maintaining compatibility with existing channel-based playback systems

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The extracted audio objects with metadata can be universally applied to various playback configurations (stereo, surround 5.1, 7.1, etc.). The same audio object data structure works across different speaker arrangements and formats, providing versatile immersive experience support without requiring format-specific processing

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

3Measurement precision

If video content is used to assist audio object extraction, then the alignment between audio and video improves, but the processing time and computational resources increase

Engineering Contradiction:
Improveaudio-video alignment accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary video object detection and metadata extraction before audio object extraction. By pre-processing the video content to identify objects, their positions, velocities, and sizes, the system prepares reference information that accelerates the subsequent audio object extraction and alignment process

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses video object metadata as feedback to guide and constrain audio object extraction. The known video object positions and characteristics provide feedback signals that help identify corresponding audio objects more quickly and accurately, reducing the computational search space and processing time

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3262640B1Video content assisted audio object extraction
Publication Date: 2020.08.05 DOLBY LABORATORIES LICENSING CORP
  • EP3262640B1 patent drawingFigure 1~2
  • EP3262640B1 patent drawingFigure 3~4
  • EP3262640B1 patent drawingFigure 5~6

AI summary

Embodiments of the present invention relate to video content assisted audio object extraction. A method of audio object extraction from channel-based audio content is disclosed. The method comprises extracting at least one video object from video content associated with the channel-based audio content, and determining information about the at least one video object. The method further comprises extracting from the channel-based audio content an audio object to be rendered as an upmixed audio signal based on the determined information. Corresponding system and computer program product are also disclosed.