Spectral Coefficient Arithmetic Coding with Quantized Context Classes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Arithmetic coding for multimedia data compression faces challenges with high encoding/decoding latency and memory capacity requirements due to the large number of contexts and corresponding probability density functions that need to be handled.
Innovation Solution
The method uses preceding spectral coefficients to determine context classes, which are then mapped to probability density functions for arithmetic encoding/decoding, employing non-uniform quantization to group similar contexts and reduce the number of functions needed, thereby decreasing latency and memory requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a large neighbourhood of preceding coefficients is used for context-based arithmetic coding, then compression efficiency is improved, but the number of contexts and probability density functions increases dramatically
Solution Approach 1:
The patent merges multiple similar contexts into a single context class. Instead of maintaining separate probability density functions for each individual context, the invention groups contexts with similar characteristics together and applies a shared probability density function to the merged group, thereby reducing the total number of functions needed.
Solution Approach 2:
The patent changes the parameter representation by using non-uniform quantization on the neighbourhood coefficients. This transforms the continuous or fine-grained context parameters into discrete quantized values, which enables grouping of similar contexts and reduces the number of distinct probability density functions required.
2Measurement precision
If many probability density functions are stored and handled for different contexts, then compression accuracy is improved, but memory capacity requirements increase
Solution Approach 1:
The patent reduces memory requirements by merging multiple probability density functions into fewer shared functions. By grouping contexts that require similar statistical models and assigning them common probability density functions, the total number of functions that need to be stored in memory is significantly reduced while maintaining adequate compression accuracy.
Solution Approach 2:
The patent makes probability density functions universal by designing them to serve multiple context classes. A single probability density function is configured to handle multiple different context situations, thereby reducing the total number of functions needed and the memory capacity required to store them.
3Productivity
If a large number of contexts are handled in arithmetic coding, then compression performance is improved, but encoding/decoding latency increases
Solution Approach 1:
The patent reduces encoding/decoding latency by merging multiple context handling operations into fewer unified operations. By grouping contexts and using shared probability density functions, the encoder and decoder need to process fewer distinct context states, which speeds up the arithmetic coding/decoding process while maintaining compression performance.
Solution Approach 2:
The patent applies non-uniform quantization to transform the context parameters into a reduced set of discrete values. This parameter transformation enables faster context identification and probability density function selection during encoding and decoding, thereby reducing processing latency while preserving the essential contextual information needed for compression.
Data Source
AI summary
The invention proposes a method and a device for arithmetic encoding of a current spectral coefficient using preceding spectral coefficients. Said preceding spectral coefficients are already encoded and both, said preceding and current spectral coefficients, are comprised in one or more quantized spectra resulting from quantizing time-frequency-transform of video, audio or speech signal sample values.Said method comprises processing the preceding spectral coefficients, using the processed preceding spectral coefficients for determining a context class being one of at least two different context classes, using the determined context class and a mapping from the at least two different context classes to at least two different probability density functions for determining the probability density function, and arithmetic encoding the current spectral coefficient based on the determined probability density function wherein processing the preceding spectral coefficients comprises non-uniformly quantizing absolutes of the preceding spectral coefficients for use in determining of the context class.