Use compressibility determination to predict the compression ratio of data.

By using a trained neural network that leverages the product of concentration and the number of zero values ​​as a threshold, and combining this with the receiver operating characteristic curve to optimize the horizontal threshold, the problem of making fast and accurate compression decisions in hardware is solved, thus improving data transmission and storage efficiency.

CN115843366BActive Publication Date: 2026-05-26HONG KONG APPLIED SCI & TECH RES INST
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HONG KONG APPLIED SCI & TECH RES INST
Filing Date
2022-09-09
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing technologies struggle to make quick and accurate decisions in hardware about whether to compress neural network data, especially under limited bandwidth conditions. Furthermore, entropy-based compression ratio prediction is complex and dependent on data characteristics.

Method used

A trained neural network is used for compression prediction. A symbol frequency table is generated through a counter, a sorter, and a shearer. The product of concentration and the number of zero values ​​is used as the input of the threshold comparator. The horizontal threshold is optimized by combining the receiver operating characteristic curve to achieve fast binary compression decision.

Benefits of technology

It enables fast and accurate decisions on whether to compress data in the hardware, reducing computational latency and resource consumption, and improving data transmission and storage efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115843366B_ABST
    Figure CN115843366B_ABST
Patent Text Reader

Abstract

The data compression analyzer can quickly make a binary decision to compress or not compress an input data block, or it can use a relatively slow neural network to predict the compression ratio of the data block through a regression model. The concentration (CV) and the number of zero-value (NZ) symbols are calculated based on the sum of squared frequencies from an unsorted symbol frequency table. A compression decision is made quickly when their product, CV*NZ, exceeds a level threshold THH. During training, CV*NZ is plotted as a function of the compression ratio C% for multiple training data blocks. Different THH test values ​​are applied to this plot to determine the true positive rate and false positive rate, and are plotted as a receiver operating characteristic (ROC) curve. The point on the ROC curve with the highest Yoden index is selected as the optimal THH for future binary decisions.
Need to check novelty before this filing date? Find Prior Art