Smart Adaptive Interpolation Filters for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video compression schemes, such as H.264/AVC and H.265/HEVC, face limitations in achieving optimal compression efficiency due to the use of static interpolation filters, which do not adapt well to varying motion patterns and image characteristics, leading to suboptimal prediction accuracy and increased bitrates for high-quality video services.
Innovation Solution
The introduction of Smart Adaptive Interpolation Filters (SAIF) that dynamically select the most suitable interpolation filters for each block of an image using a supervised learning algorithm, allowing for block-level granularity and avoiding the need for additional signaling overhead, thereby improving motion compensation prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If static interpolation filters are used in conventional video compression schemes, then device complexity is reduced and ease of operation is improved, but prediction accuracy deteriorates and bitrate increases for high-quality video services
Solution Approach 1:
The patent applies dynamics by transitioning from static interpolation filters to dynamic filter selection. A supervised learning algorithm (neural network) is introduced that adapts the interpolation filter choice based on motion characteristics and image content of each block, enabling the system to dynamically optimize prediction accuracy for varying video conditions while maintaining manageable complexity through learned patterns.
Solution Approach 2:
The patent changes the parameter of filter selection from fixed to variable. Instead of using a single static filter for all blocks, the system varies the interpolation filter parameters based on block-specific characteristics such as motion magnitude and direction, allowing optimal prediction accuracy for each block while the neural network learns the mapping from characteristics to optimal filter parameters.
2Measurement precision
If block-level granularity is implemented for filter selection, then prediction accuracy is improved, but signaling overhead increases
Solution Approach 1:
The patent applies self-service by enabling the decoder to autonomously select the appropriate interpolation filter for each block using the same supervised learning algorithm that the encoder uses. The neural network model is trained offline and deployed at both encoder and decoder, allowing the decoder to self-determine the optimal filter based on block characteristics without receiving explicit filter selection signals from the encoder, thereby eliminating the signaling overhead.
3Productivity
If conventional static interpolation filters are used, then encoding speed is maintained, but compression efficiency deteriorates for videos with varying motion patterns
Solution Approach 1:
The patent changes the interpolation filter parameters dynamically based on motion pattern characteristics. The supervised learning algorithm analyzes block-specific features such as motion magnitude and direction, then selects or adjusts filter parameters accordingly, enabling the system to adapt to varying motion patterns throughout the video sequence and significantly improving compression efficiency compared to static filter approaches.
Solution Approach 2:
The patent applies local quality by allowing different interpolation filters to be applied to different blocks based on their local characteristics. Instead of using a uniform filter across the entire video or even entire frames, each block can have its own optimized filter selection based on local motion patterns, ensuring optimal prediction accuracy and compression efficiency for each region.
Data Source
AI summary
A method of processing a first image in a first plurality of images, wherein the first image is divided into a plurality of pixel blocks, is proposed, which comprises, for a current block of the first image: selecting, in a set of a plurality of predefined interpolation filters, an interpolation filter based on a prediction of an interpolation filter determined by a supervised learning algorithm to which data related to the current block is input; and using the selected interpolation filter for calculating fractional pixel values in a second image of the plurality of images for a temporal prediction of pixels of the current block based on a reference block correlated to the current block in the second image, wherein the second image is distinct from the first image and was previously encoded according to an image encoding sequence for encoding the images of the plurality of images.