Video Encoder Mesh-Based Transform Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The High Efficiency Video Coding (HEVC) standard faces increased complexity and computational requirements due to larger transform sizes, which hinder coding efficiency and processing speed during video encoding.
Innovation Solution
The solution involves decomposing large forward transforms into multiple stages, such as using a mesh-based method, and constraining the internal bit depth at each stage to accommodate computationally efficient instruction sets, while measuring deviations and recomputing coefficients if necessary to maintain video quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If large transform sizes (16×16 and 32×32) are used in HEVC, then coding efficiency is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent applies segmentation by decomposing large transform matrices (16×16 and 32×32) into smaller block matrices. The transform process is divided into multiple stages where each stage processes smaller sub-blocks independently. This allows the large transform to be computed through a series of smaller, more manageable operations, reducing the computational burden on processing resources while maintaining the overall transform accuracy needed for coding efficiency.
2Productivity
If large transform sizes (16×16 and 32×32) are used in HEVC, then coding efficiency is improved, but processing speed decreases
Solution Approach 1:
The transform is segmented into multiple computational stages, where each stage processes a portion of the transform coefficients. This parallelizable structure allows different stages to be processed concurrently or in an optimized sequence, improving processing speed compared to a monolithic large transform computation.
Solution Approach 2:
The patent performs preliminary decomposition of the large transform into smaller block transforms before the actual coefficient computation. By pre-organizing the transform structure into manageable blocks and preparing the computational framework in advance, the actual processing requires fewer computational steps, thereby improving processing speed while maintaining coding efficiency.
3Measurement precision
If transform internal bit depth is increased, then transform accuracy is improved, but computational resource requirements increase
Solution Approach 1:
The patent applies local quality by using different bit depths for different stages of the transform computation. Rather than uniformly increasing bit depth throughout the entire transform process, the implementation uses higher precision only where necessary for maintaining accuracy, and lower precision where sufficient. This localized approach to precision management maintains transform accuracy while significantly reducing overall computational resource requirements.
Data Source
AI summary
Systems and methods for low complexity forward transforms using mesh-based calculations are described herein. One aspect of the subject matter described in the disclosure provides a video encoder comprising a memory configured to store video information. The video encoder further comprises a processor in communication with the memory. The processor is configured to decompose a transform into multiple transform stages. The processor is further configured to transform the video information using the multiple stages to determine a transform stage output at each transform stage. The processor is further configured to constrain the transform stage output at each transform stage to a predetermined bit depth. The processor is further configured to perform operations on the constrained transform output of a last stage of the multiple stages, wherein the operations are only available for use with data having the predetermined bit depth.