CCLM Parameter Inheritance for Lower-Bitrate Video Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing video coding standards, such as VVC, face challenges in efficiently signaling cross-component linear model (CCLM) related information, which is necessary for improving coding efficiency but requires additional data transmission.
Innovation Solution
A method and apparatus are introduced to derive and update model parameters for CCLM prediction, using inherited parameters from neighboring blocks and reducing the need for explicit signaling of model parameters by employing differential updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If CCLM related information is explicitly signaled to improve coding efficiency, then prediction accuracy is improved, but bit rate increases
Solution Approach 1:
The patent extracts only the necessary update information (difference values) from the full model parameters and signals only these differences rather than the complete parameters. This allows the decoder to reconstruct accurate prediction models using minimal bit rate, resolving the contradiction between prediction accuracy and bit rate consumption.
Solution Approach 2:
The patent changes the representation of model parameters from absolute values to differential updates. By signaling the difference between current and inherited parameters rather than the full parameter values, the system achieves the same prediction accuracy with significantly reduced bit rate requirements.
2Productivity
If more CCLM related information is transmitted to enhance coding efficiency, then prediction performance is improved, but data transmission requirement increases
Solution Approach 1:
The patent extracts only the essential update information needed to refine prediction models and transmits only these extracted differences. This selective transmission approach maintains coding efficiency while minimizing data transmission requirements compared to sending complete parameter sets.
Solution Approach 2:
The patent uses inheritance mechanisms where the decoder copies inherited model parameters from neighboring blocks and then applies minimal updates. This copying approach allows efficient prediction without requiring full parameter transmission, reducing data transmission while maintaining coding efficiency.
3Loss of information
If model parameters are inherited from neighboring blocks to reduce signaling, then bit rate is reduced, but parameter accuracy may deteriorate
Solution Approach 1:
The patent implements a feedback mechanism where the decoder calculates prediction models using inherited parameters and update information, then uses these models to generate predictions that can be compared with actual data. This feedback loop ensures that parameter accuracy is maintained even when using inherited parameters with minimal signaling.
Solution Approach 2:
The patent applies parameter changes by using differential updates that adjust inherited parameters to match local characteristics. This allows the system to maintain high parameter accuracy while using inheritance to reduce bit rate, as the updates compensate for any inaccuracies in the inherited base parameters.
Data Source
AI summary
A method and apparatus for video coding are disclosed. According to the method for the decoder side, the method decodes colour pictures using coding tools including one or more CCLM (Cross Colour Linear Model) related modes. The method receives encoded data associated with a current block comprising a first-colour block and a second-colour block and determines an inherited model parameter set from a previously coded block coded in a first CCLM related mode, wherein the inherited model parameter set comprises a first scaling parameter associated with the first CCLM related mode. The method further decodes the encoded data associated with the second-colour block using prediction data based on a CCLM related model associated with the inherited model parameter set.