Decoder-Side Quantization Shifting Offset Selection for Block Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently predicting quantization shifting offsets at the decoder side, leading to suboptimal image quality in reconstructed video blocks.
Innovation Solution
Implement a method for decoder side quantization shifting offset prediction by determining multiple hypotheses for potential quantization shifting offset settings, calculating cost values for each hypothesis, and selecting the hypothesis with the lowest cost to reconstruct transform coefficients, thereby improving image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple hypotheses for quantization shifting offset prediction are determined and evaluated, then image quality is improved, but computational complexity increases
Solution Approach 1:
The patent segments the quantization shifting offset prediction into multiple discrete hypotheses (e.g., different offset values like -1, 0, 1). Each hypothesis is evaluated separately using cost functions, allowing the system to find the optimal offset without exhaustive search. This segmentation transforms a continuous optimization problem into discrete, manageable candidates.
Solution Approach 2:
The patent performs preliminary action by pre-defining a set of candidate hypotheses for quantization shifting offsets before the actual reconstruction process. These hypotheses are prepared in advance based on statistical analysis or neighboring block information, so that during decoding, only these pre-selected candidates need to be evaluated, reducing the search space and computational load.
2Measurement precision
If quantization shifting offset prediction is performed at decoder side, then reconstruction accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies partial action by performing quantization shifting offset prediction only for significant transform coefficients above a certain threshold, rather than for all coefficients. This selective approach maintains reconstruction accuracy for important details while reducing processing time for less significant data, achieving a balance between accuracy and speed.
Solution Approach 2:
The patent uses feedback mechanisms where cost functions evaluate the impact of different offset hypotheses on reconstruction quality, and this evaluation feedback guides the selection of the optimal offset. The system continuously refines its predictions based on the observed performance, adjusting hypotheses to minimize reconstruction error while maintaining efficient processing.
Data Source
AI summary
An apparatus includes processing circuitry configured to receive a bitstream that comprises coded information of a current block, and determine a plurality of hypotheses for a decoder side quantization shifting offset prediction, a hypothesis in the plurality of hypotheses corresponds to a potential quantization shifting offset setting in a transform domain of the current block. The processing circuitry calculates cost values respectively associated with the plurality of hypotheses, select a specific hypothesis from the plurality of hypotheses according to the cost values, determine one or more quantization shifting offsets for transform coefficients in the transform domain based on the specific hypothesis, reconstruct the transform coefficients based on the one or more quantization shifting offsets, calculate residuals in a spatial domain of the current block based on the transform coefficients in the transform domain, and reconstruct the current block according to the residuals in the spatial domain.