Color Component Prediction Using Reference Pixel Subset Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing color prediction models in video coding and decoding require a large number of samples for model construction, leading to high calculation complexity and memory bandwidth, and are prone to inaccuracies due to exceptional samples.
Innovation Solution
Reduce the number of pixels in the reference pixel set by selecting a subset of candidate pixels, removing unimportant or exceptional reference pixels to improve the accuracy and efficiency of the prediction model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a large number of samples are used to construct the prediction model, then the model can cover more data points, but the calculation complexity and memory bandwidth increase significantly
Solution Approach 1:
The patent extracts only the necessary and representative reference pixels from the full reference pixel set to construct the prediction model. By selecting a subset of candidate pixels that are most relevant for prediction (e.g., those with higher correlation or representativeness), the method reduces the number of samples needed while maintaining model accuracy, thereby lowering calculation complexity and memory bandwidth requirements.
Solution Approach 2:
The patent applies local quality by treating different reference pixels differently based on their importance or representativeness. Instead of uniformly using all reference pixels, the method identifies and weights or selects specific pixels that locally best represent the color characteristics, improving model accuracy with fewer samples and reducing overall computational burden.
2Reliability
If a large number of samples are used to construct the prediction model, then more data is available for modeling, but the memory bandwidth requirement increases
Solution Approach 1:
The patent extracts only the essential reference pixels needed for accurate color component prediction. By identifying and selecting a subset of candidate pixels from the reference pixel set based on their representativeness or correlation strength, the method reduces the volume of data that needs to be stored and transmitted, thereby decreasing memory bandwidth consumption while preserving prediction model accuracy.
3Reliability
If all reference pixels are used in the prediction model, then comprehensive color information is captured, but exceptional samples reduce prediction accuracy
Solution Approach 1:
The patent removes exceptional or outlier samples from the reference pixel set that would otherwise degrade prediction accuracy. By extracting and selecting only the high-quality, representative candidate pixels that truly reflect the color characteristics, the method eliminates the negative impact of exceptional samples while maintaining comprehensive color information representation.
Solution Approach 2:
The patent applies local quality by differentiating between high-quality and low-quality reference pixels. Instead of treating all pixels uniformly, the method identifies pixels with superior representativeness and uses them preferentially in the prediction model, thereby improving prediction accuracy by focusing on the most reliable local color information.
4Reliability
If more reference pixels are selected for the prediction model, then the model covers more variations, but the calculation time increases
Solution Approach 1:
The patent extracts the most critical reference pixels that capture the essential color variations needed for robust prediction. By selecting a compact subset of candidate pixels that are most representative of color variations, the method achieves model robustness with minimal calculation time, avoiding the computational overhead of processing all reference pixels.
Data Source
AI summary
Provided are a method for predicting a colour component, an encoder and a decoder. The method includes: a first reference sample set of a colour component to be predicted of a current block is determined; a reference sample subset is determined from the first reference sample set, where the reference sample subset includes one or more candidate samples selected from the first reference sample set; and a model parameter of a prediction model is calculated according to positions of reference samples in the reference sample subset, where the prediction model is configured to perform prediction processing on the colour component to be predicted of the current block.