Fast Affine Motion Estimation for VVC using Machine-Learning-Based Early Search Termination

Duarte, Adson; Gonçalves, Paulo Romeu; Agostini, Luciano; Corrêa, Guilherme; Porto, Marcelo; Palomino, Daniel

doi:10.1109/iscas48785.2022.9937973

Cited by 9 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To do this, each element below the pivot row is recalculated according to Eq. (11), where col and row are the columns and line from each element inside the matrix, and the PivotCol and PivotRow are the columns and line from the pivot inside the matrix.…”

Section: Linear Equation System Solvingmentioning

confidence: 99%

“…Substitution: Starting from the last equation (bottom row), solve for the variables one by one, working your way up to the first equation. (11) In the next section, a software analysis of the affine Inter-Mode will be presented, focusing on the affine MCs performed in the iterative algorithm of the affine Inter-Mode, which is where the linear equation must be solved to generate a new set of MVs.…”

Section: Linear Equation System Solvingmentioning

confidence: 99%

“…However, this improvement comes at the cost of a substantial increase in complexity. For example, the affine prediction tool represents 43% of the total inter-frame prediction complexity on VTM version 9 [11].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

High-Throughput Hardware Design for Linear Equation System Solving of VVC Affine Prediction

Maass,

Muñoz,

Perleberg

et al. 2023

JICS

View full text Add to dashboard Cite

The affine prediction is the main novelty in the inter-frame prediction of Versatile Video Coding (VVC) standard. However, implementing the affine prediction requires a huge computational effort that makes mandatory the use of dedicated hardware accelerators to achieve real-time processing and meet the constraints of area and power dissipation for mobile devices. In light of this, this work presents different approaches for a hardware design dedicated to solving the linear equation system, which is essential for refiningthe affine Motion Vector. Synthesis results show that the High- Throughput architecture, which explores the parallelism of internal operations, can reach an accuracy of 99.99%, requiring 333.9kgates to be implemented and presenting a power dissipation of 120.4mW when running at 540MHz, the operational frequency required to process 60 frames per second of HD 1080p videos.

show abstract

Section: Linear Equation System Solvingmentioning

confidence: 99%

Section: Linear Equation System Solvingmentioning

confidence: 99%