V-DMC Mesh Motion Coding With Weighted Vector Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mesh compression techniques struggle with efficiently encoding and decoding complex 3D mesh sequences, leading to large file sizes and high bandwidth requirements, which affect the visual quality of 3D video and virtual reality content, especially in dynamic scenes.
Innovation Solution
Implementing motion vector prediction techniques that utilize weighted averaging of neighboring vertices to predict the motion vector for a current vertex, incorporating historical information and prioritizing spatially closer vertices, thereby enhancing accuracy and reducing the data needed for transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If mesh compression techniques are used to reduce file sizes, then the quantity of data is reduced, but the manufacturing precision of motion vector encoding deteriorates
Solution Approach 1:
The patent introduces motion vector predictors as intermediary elements that mediate between the current motion vector and the encoding process. These predictors are derived from neighboring vertices and temporal references, serving as intermediate representations that reduce the precision requirements of direct motion vector encoding while maintaining overall accuracy through the prediction mechanism.
Solution Approach 2:
The patent changes the parameter representation by transitioning from encoding absolute motion vectors to encoding prediction residuals. This parameter transformation allows for more efficient compression by exploiting the statistical properties of residuals, which have smaller magnitudes and better probability distributions for compression compared to raw motion vectors.
2Device complexity
If simple averaging of motion vectors is used, then the device complexity is reduced, but the measurement precision of motion vector prediction deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the prediction process for different spatial locations. Different weighting schemes are applied based on the local characteristics of the mesh, with closer vertices receiving higher weights. This localized approach improves prediction accuracy in regions with varying motion patterns without uniformly increasing complexity across the entire mesh.
Solution Approach 2:
The patent introduces dynamic adaptivity in the prediction process by allowing the prediction mechanism to adjust based on temporal and spatial context. The selection of predictors and weighting factors becomes dynamic rather than static, enabling the system to adapt to different motion scenarios while maintaining manageable complexity through context-based decision making.
3Measurement precision
If more motion vector data is transmitted to improve prediction accuracy, then the measurement precision is improved, but the bandwidth requirements increase
Solution Approach 1:
The patent performs preliminary action by pre-computing motion vector predictors at the encoder side using available reference information and neighboring vertices. This preliminary computation reduces the amount of data that needs to be transmitted, as only the residuals between actual motion vectors and predictors need to be encoded and transmitted, rather than transmitting complete motion vector information for all vertices.
Solution Approach 2:
The patent implements feedback mechanisms where the encoder and decoder use the same prediction algorithms to generate identical predictors, enabling the decoder to reconstruct motion vectors accurately from transmitted residuals. This feedback-based approach ensures that the transmitted data volume is minimized while maintaining prediction accuracy through synchronized encoder-decoder prediction processes.
Data Source
AI summary
A method of encoding or decoding mesh data includes: for a current vertex of mesh vertices of the mesh data, determining a motion vector predictor based on respective weighted averages of respective motion vectors in a candidate list for the current vertex; and encoding or decoding the current vertex based on the motion vector predictor.