Adaptive Loop Filter Memory Reduction via Temporal Layer Arrays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards, such as JEM 7, require significant memory for temporal prediction in adaptive loop filtering, which can be inefficient and not scalable, especially in hierarchical frame structures.
Innovation Solution
Implementing a one-dimensional array to store adaptive loop filter (ALF) parameters along with temporal layer identifiers, reducing memory usage while maintaining temporal scalability by storing and filtering video data blocks using these parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If temporal prediction in adaptive loop filtering is implemented using conventional methods (e.g., JEM 7), then compression performance is maintained, but memory consumption increases significantly
Solution Approach 1:
The patent segments the temporal prediction process by dividing reference pictures into multiple temporal layers and selectively applying ALF parameters from different layers based on temporal distance. This segmentation allows the system to use fewer memory resources while maintaining compression performance for different picture types (I, P, B slices) through layered parameter selection.
Solution Approach 2:
The patent changes the parameters of ALF filtering by selecting different filter strengths and configurations based on the temporal layer ID and picture type. By dynamically adjusting filtering parameters according to temporal distance and slice type, the system achieves effective compression with reduced memory requirements compared to uniform strong filtering.
2Productivity
If adaptive loop filtering with temporal prediction is implemented, then compression efficiency improves, but device complexity increases
Solution Approach 1:
The patent introduces dynamic adaptation of ALF parameters based on temporal layer identification and picture type (I, P, B slices). The filtering behavior dynamically adjusts by selecting parameters from appropriate temporal layers, enabling efficient compression without requiring complex uniform filtering across all picture types.
Solution Approach 2:
The patent applies different ALF parameter sets to different local regions (I slices, P slices, B slices) based on their temporal layer characteristics. Each slice type receives optimized filtering parameters suited to its temporal prediction needs, improving compression efficiency while avoiding the complexity of a fully adaptive system.
3Reliability
If strong adaptive loop filtering is applied to all picture types, then compression performance is maximized, but processing overhead and memory usage increase
Solution Approach 1:
The patent applies partial filtering action by selectively enabling ALF parameters only where needed based on temporal layer ID and picture type. Instead of applying strong filtering uniformly to all slices, the system applies filtering partially to I, P, and B slices according to their temporal characteristics, reducing processing overhead while maintaining necessary compression performance.
Data Source
AI summary
A device for decoding video data includes a memory configured to store video data and one or more processors implemented in circuitry and configured to store sets of adaptive loop filter (ALF) parameters in a one-dimensional array in the memory, the one-dimensional array having a predefined size of N memory elements, N being a positive integer value, wherein the one or more processors are configured to store, in one or more of the memory elements of the array, both corresponding ALF parameters and a temporal layer identifier (ID) value indicating a temporal layer from which the corresponding ALF parameters are to be estimated; decode one or more blocks of the video data; and filter the one or more blocks using the ALF parameters of the one-dimensional array. The device may further encode the one or more blocks prior to decoding the one or more blocks.