Inter-Prediction Mode Binarization for Small-PU Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high memory bandwidth requirements for motion compensation in video coding, particularly in HEVC, pose a bottleneck for embedded decoder implementations, especially with smaller prediction unit sizes, which are exacerbated by bi-predicted 4×8 or 8×4 prediction blocks.
Innovation Solution
Restricting inter-coded prediction units of small block sizes in bi-predicted slices to uni-prediction mode during encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bi-prediction is used for small prediction units (4×8 or 8×4 blocks), then prediction accuracy is improved, but memory bandwidth consumption increases significantly
Solution Approach 1:
The patent applies local quality by differentiating the prediction mode based on prediction unit size. Small PUs (4×8 or 8×4) are restricted to uni-prediction mode while larger PUs can use bi-prediction. This localized restriction reduces memory bandwidth consumption for small blocks where bi-prediction provides minimal benefit, while preserving prediction accuracy for larger blocks where it is more beneficial.
Solution Approach 2:
The patent changes the prediction mode parameter based on PU size. By modifying the prediction mode from bi-prediction to uni-prediction for small PUs, the patent reduces the memory bandwidth requirements. This parameter change is implemented through syntax elements in the bitstream that indicate whether bi-prediction is allowed for small PUs, enabling dynamic adjustment of prediction behavior.
2Productivity
If HEVC uses larger block sizes and 8-tap filters for motion compensation, then coding efficiency is improved, but memory bandwidth requirements increase by about 50%
Solution Approach 1:
The patent applies local quality by restricting bi-prediction only for small prediction units (4×8 or 8×4 blocks) while allowing it for larger PUs. This selective approach reduces memory bandwidth consumption for small blocks where the benefit of bi-prediction is minimal, while preserving coding efficiency for larger blocks where bi-prediction provides significant improvement.
Solution Approach 2:
The patent implements partial action by not completely disabling bi-prediction for all PUs, but only for small ones (4×8 or 8×4). This partial restriction reduces memory bandwidth consumption while maintaining coding efficiency for larger PUs where bi-prediction is still permitted and beneficial.
3Quantity of substance
If inter-prediction of 4×4 prediction blocks is disabled, then memory bandwidth requirements are reduced, but prediction flexibility is lost
Solution Approach 1:
The patent applies local quality by differentiating between small PUs (4×8 or 8×4) and larger PUs. Bi-prediction is restricted only for small PUs to reduce memory bandwidth consumption, while larger PUs retain bi-prediction capability to maintain prediction flexibility. This localized approach balances memory efficiency with prediction versatility.
Solution Approach 2:
The patent implements dynamic behavior by allowing the prediction mode to vary based on PU size. The restriction on bi-prediction for small PUs is not absolute but can be controlled through syntax elements in the bitstream, enabling dynamic adjustment of prediction flexibility based on the specific coding context and PU dimensions.
Data Source
AI summary
Motion compensation requires a significant amount of memory bandwidth, especially for smaller prediction unit sizes. The worst case bandwidth requirements can occur when bi-predicted 4×8 or 8×4 PUs are used. To reduce the memory bandwidth requirements for such smaller PUs, methods are provided for restricting inter-coded PUs of small block sizes to be coded only in a uni-predictive mode, i.e., forward prediction or backward prediction. More specifically, PUs of specified restricted sizes in bi-predicted slices (B slices) are forced to be uni-predicted.