Inter Prediction Weighted Signaling Overhead Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution, high-quality images and videos, particularly in applications like virtual reality and augmented reality, necessitates a more efficient compression technique to reduce transmission and storage costs while effectively handling diverse image/video characteristics.
Innovation Solution
A method and apparatus for enhancing video/image coding efficiency through improved inter prediction, including efficient signaling of a prediction weighted table syntax and reducing signaling overhead by parsing flags related to weighted prediction, thereby optimizing the use of bits for weighted prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high resolution and high quality image/video are transmitted or stored using existing methods, then image quality is maintained, but transmission and storage costs increase significantly
Solution Approach 1:
The patent applies parameter changes by modifying the precision of motion vector representation. Instead of using full-precision motion vectors for all blocks, the invention uses reduced-precision motion vectors (e.g., integer pixel precision instead of sub-pixel precision) for certain prediction blocks, particularly when the block can be adequately predicted with lower precision. This reduces the number of bits required to encode motion information while maintaining acceptable prediction quality, thereby reducing transmission and storage costs.
2Productivity
If conventional inter prediction methods are used, then encoding simplicity is maintained, but coding efficiency decreases for high resolution videos
Solution Approach 1:
The patent applies segmentation by dividing the current picture into multiple prediction blocks and applying different prediction methods to different blocks. Specifically, the invention segments the prediction process into: (1) blocks that use reduced-precision motion vectors, (2) blocks that use full-precision motion vectors, and (3) blocks that use different reference picture lists. This selective application of different prediction strategies to different blocks improves overall coding efficiency while managing complexity through localized decision-making.
Solution Approach 2:
The patent applies dynamics by making the prediction method adaptive rather than static. The invention dynamically selects between reduced-precision and full-precision motion vectors based on block characteristics, such as motion complexity, block size, and predicted prediction quality. This dynamic adaptation allows the encoding process to optimize for each block individually, improving overall coding efficiency without uniformly increasing complexity across the entire picture.
3Measurement precision
If full precision motion vectors are used for all prediction blocks, then prediction accuracy is maximized, but the number of bits required for encoding increases
Solution Approach 1:
The patent applies parameter changes by selectively changing the precision parameter of motion vectors based on block characteristics. Instead of using a fixed high precision for all blocks, the invention changes the precision parameter dynamically: using reduced precision (e.g., integer pixel) for blocks where high precision is not critical, and full precision (e.g., sub-pixel) for blocks where it is necessary. This selective parameter change reduces the total number of bits required to encode motion information while maintaining prediction accuracy where needed.
4Measurement precision
If weighted prediction is applied to all blocks, then prediction quality improves, but signaling overhead increases
Solution Approach 1:
The patent applies segmentation by dividing blocks into categories based on whether weighted prediction should be applied. The invention segments the prediction process to apply weighted prediction selectively to certain blocks (e.g., blocks with significant motion or specific characteristics) while using standard prediction for other blocks. This segmentation reduces the signaling overhead associated with weighted prediction parameters while maintaining prediction quality for blocks that benefit most from it.
Data Source
AI summary
A video decoding method performed by a video decoding apparatus, according to the present document, comprises the steps of: parsing, from a bitstream, flags associated with weighted prediction; parsing a prediction weighted table syntax from a picture header of the bitstream on the basis of the flags; generating prediction samples regarding a current block in a current picture on the basis of syntax elements in the prediction weighted table syntax; and generating reconstructed samples on the basis of the prediction samples, wherein the flags comprise a first flag associated with whether the weighted prediction is applied to slices referring to a picture parameter set of the bitstream, and a second flag associated with whether information related to the weighted prediction is present in the picture header, and the second flag may be parsed from the picture parameter set on the basis of the first flag.