Dual Probability Estimation for Adaptive Multi-Symbol Entropy Coding

Solution Overview

Problem

Existing video compression techniques face challenges in accurately estimating cumulative probability distributions, leading to suboptimal coding efficiency due to either fast or slow adaptation rates.

Innovation Solution

The use of a mixture of two or more adaptive cumulative probability estimates, each updated with different adaptation parameters, to balance the pros and cons of fast and slow adaptation, thereby improving coding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a single adaptive cumulative probability estimate is updated with a fast adaptation rate, then the system can quickly adapt to probability changes, but the precision deteriorates when probabilities are stable

Engineering Contradiction:

Improveadaptation speedVSAvoidprobability estimation precision

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments the single probability estimation function into two separate estimators: a first adaptive cumulative probability estimator with fast adaptation rate and a second adaptive cumulative probability estimator with slow adaptation rate. Each estimator handles different scenarios, allowing the system to switch between them based on whether probabilities are changing or stable, thus resolving the contradiction between adaptation speed and precision.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces dynamic switching between two estimators based on the detected state of probability distribution. When a symbol probability change is detected, the system dynamically switches to the first estimator for quick adaptation; when no change is detected, it switches to the second estimator for precise estimation. This dynamic behavior allows the system to optimize both adaptation speed and precision at different times.

Inventive Principle:

Principle #15Dynamics

2Measurement precision

If a single adaptive cumulative probability estimate is updated with a slow adaptation rate, then the precision is maintained when probabilities are stable, but the adaptation speed deteriorates when probabilities change

Engineering Contradiction:

Improveprobability estimation precisionVSAvoidadaptation speed

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent segments the single probability estimation function into two separate estimators: a first adaptive cumulative probability estimator with fast adaptation rate and a second adaptive cumulative probability estimator with slow adaptation rate. Each estimator handles different scenarios, allowing the system to switch between them based on whether probabilities are changing or stable, thus resolving the contradiction between adaptation speed and precision.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces dynamic switching between two estimators based on the detected state of probability distribution. When a symbol probability change is detected, the system dynamically switches to the first estimator for quick adaptation; when no change is detected, it switches to the second estimator for precise estimation. This dynamic behavior allows the system to optimize both adaptation speed and precision at different times.

Inventive Principle:

Principle #15Dynamics

3Ease of manufacture

If video compression uses traditional entropy coding methods, then the implementation is simple, but the coding efficiency deteriorates due to inaccurate probability estimation

Engineering Contradiction:

Improveimplementation simplicityVSAvoidcoding efficiency

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The patent segments the probability estimation into two specialized estimators that can be selectively applied. This segmentation allows the system to maintain relatively simple implementation while significantly improving coding efficiency by using the appropriate estimator based on the current probability distribution state, resolving the contradiction between simplicity and efficiency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes the adaptation rate parameter dynamically by selecting between two different estimators with different adaptation rates. This parameter change allows the system to improve coding efficiency by using fast adaptation when probabilities change and slow adaptation when probabilities are stable, while keeping the overall implementation structure relatively simple.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentEP4550784A1Probability estimation in multi-symbol entropy coding with two estimators

Publication Date: 2025.05.07 INTEL CORP

Figure 1
Figure 2
Figure 3

AI summary

Entropy coding, such as multi-symbol arithmetic coding, is used in video compression to encode data into a compressed bit stream for transmission. Some entropy coding techniques are adaptive, meaning that the probability distribution is updated on the fly, based on the data. Accuracy of cumulative probability estimation in adaptive multi-symbol arithmetic coding can impact coding efficiency. To address the issue, a mixture of two or more adaptive cumulative probability estimations computed using two or more adaptation parameters can be used in place of a single cumulative probability estimate. The two or more adaptation parameters can be unique for a context model. A divergence in the adaptive cumulative probability estimations may signal a sudden change in the probability of a symbol. The divergence may trigger a reset of one or more adaptive cumulative probability estimations.