Entropy Encoding with Mismatch Codewords for Unseen Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing entropy encoding methods fail to efficiently compress data that has not been previously processed, leading to inefficient compression of previously unseen data.
Innovation Solution
Implement mismatch probability estimation to calculate the frequency of occurrence of data sourceblocks and insert a mismatch codeword into the codebook, using a secondary encoding process when a mismatch occurs during encoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional entropy encoding methods are used, then compression is efficient for previously seen data, but compression efficiency deteriorates for previously unseen data
Solution Approach 1:
The patent performs preliminary actions by pre-calculating mismatch probabilities during a training phase before actual encoding occurs. The system prepares probability estimates for unseen data patterns in advance, allowing the encoder to handle novel data efficiently without requiring extensive retraining or adaptation during the encoding process itself.
Solution Approach 2:
The patent dynamically changes encoding parameters by adjusting codeword assignments based on calculated mismatch probabilities. The system modifies the probability distribution parameters used in entropy encoding to account for unseen data, thereby adapting the compression scheme to handle both familiar and novel data patterns effectively.
2Measurement precision
If codebook size is increased to cover more data patterns, then compression accuracy improves, but device complexity increases
Solution Approach 1:
The patent changes the parameter representation by using probability estimates rather than exhaustive codebook entries. Instead of increasing codebook size to cover all possible patterns, the system modifies the probability parameters to reflect the likelihood of unseen patterns, achieving accurate compression with a compact codebook.
Solution Approach 2:
The patent introduces mismatch probability estimates as an intermediary mechanism between the codebook and unseen data. This intermediary allows the system to handle patterns not explicitly represented in the codebook by using probability-based predictions, thereby maintaining compression accuracy without expanding the codebook.
Data Source
AI summary
Techniques for compressing data that use mismatch probability estimation to improve entropy encoding methods to account for, and efficiently handle, previously unseen data in data to be compressed. A mismatch probability estimate is calculated comprising an estimated frequency at which any given data sourceblock received during encoding will not have a codeword in the codebook. Entropy encoding is used to generate codebooks comprising codewords for data sourceblocks based on the frequency of occurrence of each sourceblock. A “mismatch codeword” is inserted into the codebook based on the mismatch probability estimate to represent those cases when a block of data to be encoded does not have a codeword in the codebook.