Local Entropy Encoding for Block-Based Data Compression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing data compression systems are inefficient in reducing memory requirements and communication bandwidth without compromising data quality and authenticity.

Innovation Solution

The encoding system generates a compressed representation of data by partitioning it into code symbol subsets and entropy encoding each subset using a dictionary of code symbol probability distributions or custom distributions learned from the data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If conventional data compression systems are used, then data can be stored and transmitted, but memory requirements and communication bandwidth are not reduced efficiently

Engineering Contradiction:
Improvememory requirementsVSAvoidcompression efficiency
Core Design Contradiction:
Quantity of substanceVSProductivity

Solution Approach 1:

The data stream is divided into multiple blocks, and each block is processed independently using separate probability distributions. This segmentation allows the system to adapt to local statistical variations in different parts of the data, improving compression efficiency while reducing the memory buffer required to maintain context across the entire data stream.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system uses context-adaptive probability distributions that are specific to each block of data rather than a single global distribution. By learning and applying local statistical characteristics for each block, the encoder achieves better compression ratios without requiring large memory buffers to capture long-range dependencies.

Inventive Principle:
Principle #3Local quality

2Quantity of substance

If conventional data compression systems are used, then data can be stored and transmitted, but communication bandwidth is not reduced efficiently

Engineering Contradiction:
Improvecommunication bandwidthVSAvoidcompression efficiency
Core Design Contradiction:
Quantity of substanceVSProductivity

Solution Approach 1:

By processing data in blocks with independent probability models, the system achieves efficient compression that reduces the number of bits required for transmission. The block-based approach captures local redundancies effectively, minimizing the communication bandwidth needed while maintaining compression performance.

Inventive Principle:
Principle #1Segmentation

3Productivity

If compression is increased to reduce memory and bandwidth, then efficiency improves, but data quality and authenticity may be compromised

Engineering Contradiction:
Improvecompression efficiencyVSAvoiddata quality
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system applies compression selectively at the block level, using full entropy coding where appropriate while maintaining the ability to preserve exact data representation when needed. This partial application of aggressive compression techniques maintains data authenticity while achieving high overall compression efficiency.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3740912B1Data compression by local entropy encoding
Publication Date: 2025.03.05 GOOGLE LLC
  • EP3740912B1 patent drawingFigure 1
  • EP3740912B1 patent drawingFigure 2
  • EP3740912B1 patent drawingFigure 3

AI summary

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, an encoder neural network processes data to generate an output including a representation of the data as an ordered collection of code symbols. The ordered collection of code symbols is entropy encoded using one or more code symbol probability distributions. A compressed representation of the data is determined based on the entropy encoded representation of the collection of code symbols and data indicating the code symbol probability distributions used to entropy encode the collection of code symbols. In another aspect, a compressed representation of the data is decoded to determine the collection of code symbols representing the data. A reconstruction of the data is determined by processing the collection of code symbols by a decoder neural network.