Combined Inter-Intra Prediction for Lower Video Coding Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing video coding technologies, such as VVC, face challenges in efficiently combining inter and intra prediction methods due to high computational complexity, interdependency on neighboring block reconstruction, and non-unified treatment of prediction modes, which affect encoding/decoding throughput and memory bandwidth.
Innovation Solution
Simplify the combined inter and intra prediction (CIIP) by disabling bi-directional optical flow, converting bi-prediction to uni-prediction for memory efficiency, and harmonizing intra mode treatments for unified MPM candidate lists, reducing computational complexity and memory bandwidth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If combined inter and intra prediction mode is applied to a current coding block, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the prediction process by first generating separate inter prediction and intra prediction results, then selectively combining them only when beneficial. This segmentation allows the system to maintain high prediction accuracy while avoiding unnecessary computational complexity in cases where simple inter or intra prediction suffices.
Solution Approach 2:
The patent applies partial action by implementing combined prediction selectively rather than universally. The mode selection mechanism determines whether to apply the computationally intensive combined prediction or use simpler inter/intra prediction alone, thus achieving partial application of the complex method only where needed to improve accuracy.
2Measurement precision
If combined inter and intra prediction mode is applied to a current coding block, then prediction accuracy is improved, but encoding/decoding throughput decreases
Solution Approach 1:
The patent divides the prediction process into distinct stages: inter prediction generation, intra prediction generation, and optional combination. This segmentation enables the encoder/decoder to efficiently process blocks using simple prediction when applicable, maintaining high throughput while still achieving high accuracy when combined prediction is selected.
Solution Approach 2:
By applying combined prediction only partially (selectively based on mode selection) rather than to all coding blocks, the patent maintains high encoding/decoding throughput for the majority of blocks while achieving improved accuracy for specific blocks where the complex prediction provides benefit.
3Measurement precision
If bi-directional optical flow is enabled in combined prediction, then prediction accuracy is improved, but memory bandwidth increases
Solution Approach 1:
The patent extracts and disables the bi-directional optical flow component from the combined prediction process. This removal eliminates the excessive memory bandwidth consumption associated with BDOF while retaining the core combined prediction functionality, thus achieving a balance between accuracy improvement and memory efficiency.
4Adaptability or versatility
If non-unified treatment of prediction modes is used, then flexibility is improved, but device complexity increases
Solution Approach 1:
The patent implements a unified prediction mode treatment where the same combined prediction framework handles both inter and intra prediction modes. This universal approach maintains flexibility in adapting to different prediction scenarios while reducing device complexity by eliminating the need for separate, specialized processing paths for different mode types.
Data Source
AI summary
A method for video coding is provided. The method includes: determining that a combined inter and intra prediction mode is applied to a current coding block; generating an inter prediction of the current coding block based on an inter prediction mode; generating an intra prediction of the current coding block based on an intra prediction mode; generating a final prediction of the current coding block based on the inter prediction and the intra prediction; and determining that a prediction mode of the current coding block is treated as one unified mode when constructing a most probable mode (MPM) list of a neighboring coding block, the unified mode being predefined as an inter mode, or being predefined as an intra mode.