Asymmetric Probability Model Update for Video Entropy Coding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing video compression techniques face limitations in reducing data size while maintaining precision, as they often use a fixed number of bits for both probability model storage and entropy coding, leading to suboptimal compression performance due to reduced accuracy from lower bit precision during entropy decoding.

Innovation Solution

Implementing asymmetric probability model updating and entropy coding, where different bit precisions are used for probability model storage and entropy decoding, allowing for more precise model updates while using fewer bits for entropy coding, thereby improving compression efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a fixed number of bits is used for both probability model storage and entropy coding, then device complexity is reduced, but compression performance deteriorates due to reduced accuracy from lower bit precision during entropy decoding

Engineering Contradiction:

Improveprobability model precisionVSAvoidbit precision management complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies asymmetry by using different bit precisions for probability model storage and entropy coding operations. Specifically, the probability model is maintained with higher precision (first bit precision) while entropy decoding uses lower precision (second bit precision), creating an asymmetric precision architecture that optimizes both accuracy and efficiency

Inventive Principle:

Principle #4Asymmetry

Solution Approach 2:

The patent segments the bit precision requirement into two distinct components: one for probability model maintenance and another for entropy coding operations. This segmentation allows each component to operate at its optimal precision level independently, with the probability model using first bit precision and entropy decoding using second bit precision

Inventive Principle:

Principle #1Segmentation

2Productivity

If fewer bits are used for entropy decoding, then compression throughput is improved, but measurement precision of the probability model deteriorates

Engineering Contradiction:

Improvecompression throughputVSAvoidprobability model accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by maintaining the high-precision probability model in advance before entropy decoding operations. The probability model is updated and maintained with first bit precision prior to being used in entropy decoding, ensuring accuracy is preserved before the lower-precision decoding process

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent resolves this contradiction through asymmetric precision allocation where the probability model storage uses first bit precision while entropy decoding uses second bit precision. This asymmetry allows the model to maintain high accuracy while the decoding process operates efficiently with fewer bits

Inventive Principle:

Principle #4Asymmetry

3Measurement precision

If high bit precision is used for entropy decoding, then measurement precision is maintained, but compression performance deteriorates due to increased data size

Engineering Contradiction:

Improvedecoding accuracyVSAvoidbit rate

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies local quality by assigning different precision levels to different functional components. The probability model maintenance uses high precision (first bit precision) where accuracy is critical, while entropy decoding uses lower precision (second bit precision) where moderate accuracy suffices, optimizing the overall bit rate

Inventive Principle:

Principle #3Local quality

4Productivity

If asymmetric bit precision is used for probability model and entropy coding, then compression performance is improved, but device complexity increases due to dual precision management

Engineering Contradiction:

Improvecompression efficiencyVSAvoiddual precision management complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent applies dynamics by making the bit precision selection adaptive rather than fixed. The system dynamically determines the relationship between first bit precision and second bit precision based on probability thresholds, allowing the precision configuration to adapt to the statistical characteristics of the data being encoded

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS11218737B2Asymmetric probability model update and entropy coding precision

Publication Date: 2022.01.04 GOOGLE LLC

AI summary

Asymmetric probability model updating and entropy coding includes using different numbers of bits for storing probabilities of a probability model and for entropy coding symbols using that probability model. The probabilities of a probability model are updated according to values of syntax elements decoded from a bitstream. The probabilities are associated with possible values of the syntax elements and are stored using a first bit precision. Based on the updated probabilities, a second bit precision to use to entropy decode the syntax elements is determined. The second bit precision is less than the first bit precision. The syntax elements are then entropy decoded using the second bit precision, such as to produce quantized transform coefficients, which may be further processed and output to an output video stream. Using the first bit precision to entropy decode the syntax elements results in a lower compression throughput than using the second bit precision.