Dynamic Mesh Data Layout for Single-Bitstream 2D Codec Processing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing immersive video coding systems require multiple video bitstreams, which can hinder deployment in systems that handle only one bitstream, reducing throughput and efficiency.

Innovation Solution

Arrange displacement vectors and attribute maps of 3D visual media data into a single picture for processing by a 2D codec, aligning them to a common bitdepth and color format, and indicate location and size information in the bitstream for independent decoding.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple video bitstreams are used for immersive video coding, then coding efficiency and adaptability are improved, but system complexity and deployment difficulty increase

Engineering Contradiction:

Improvecoding efficiencyVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent combines displacement vectors and attribute maps into a single picture structure that can be processed by conventional 2D video codecs. This merging allows immersive video data to be encoded using standard codecs rather than requiring multiple specialized bitstreams, thereby reducing system complexity while maintaining coding efficiency through unified processing

Inventive Principle:

Principle #5Merging (Combining)

2Adaptability or versatility

If multiple video bitstreams are used for immersive video coding, then data representation flexibility is improved, but throughput and processing efficiency deteriorate

Engineering Contradiction:

Improvedata representation flexibilityVSAvoidthroughput

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent creates a universal picture structure that can be processed by any conventional 2D video codec, making the immersive video coding system compatible with existing infrastructure. This multi-functionality allows the same encoding framework to handle both displacement vectors and attribute maps, improving throughput by eliminating the need for separate processing pipelines while maintaining data representation flexibility

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Ease of operation

If displacement vectors and attribute maps are arranged in different pictures, then independent decoding capability is improved, but processing complexity and bandwidth usage increase

Engineering Contradiction:

Improveindependent decoding capabilityVSAvoidprocessing complexity

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The patent segments the single picture into different regions or slices, where displacement vectors and attribute maps are arranged in different parts. This segmentation allows independent decoding of each part while maintaining overall integration, thereby enabling independent decoding capability without significantly increasing processing complexity compared to unified processing

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS20250330627A1Data Arrangement For Dynamic Mesh Coding

Publication Date: 2025.10.23 BYTEDANCE INC

AI summary

A mechanism for processing video data is disclosed. A determination is made to arrange displacement vectors and an attribute map of three-dimensional (3D) visual media data into a single picture for processing by a two-dimensional (2D) video codec. A conversion is performed between a visual media data and a bitstream based on the displacement vectors and the attribute map.