Feature Map Decoding Using Peak Probability Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image compression technologies face challenges in ensuring image quality while improving compression efficiency, particularly in meeting the increasing demands of multimedia applications.
Innovation Solution
A feature map encoding and decoding method that determines feature elements based on a peak probability rather than a fixed probability, allowing for more accurate decoding and reduced encoding and decoding complexity by selectively performing entropy encoding on feature elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional lossy image compression technologies are used to improve compression efficiency, then compression ratio is improved, but image quality deteriorates
Solution Approach 1:
The patent changes the parameter from fixed probability to peak probability in the probability estimation result, allowing dynamic adjustment of decoding accuracy based on the actual distribution characteristics of feature elements, thereby improving image quality while maintaining compression efficiency
Solution Approach 2:
The patent introduces dynamic threshold adjustment based on peak probability values, making the decoding process adaptive to different feature element distributions rather than using static thresholds, which improves both compression efficiency and image quality
2Device complexity
If fixed probability threshold is used for determining feature elements, then encoding and decoding process is simplified, but decoding accuracy deteriorates
Solution Approach 1:
The patent performs probability estimation and identifies peak probability values in advance during the encoding process, storing this information for use during decoding. This preliminary action allows the decoder to use accurate peak probability values without increasing decoding complexity
Solution Approach 2:
The patent copies the peak probability information from the encoder to the decoder through the bitstream, allowing the decoder to use the same accurate probability values without performing complex calculations, thus maintaining simplicity while improving accuracy
Data Source
AI summary
This application provides a feature map encoding and decoding method and an apparatus, and relates to the field of artificial intelligence (AI)-based data encoding and decoding technologies. The feature map decoding method includes: obtaining a bitstream of a to-be-decoded feature map, where the to-be-decoded feature map includes a plurality of feature elements; obtaining a first probability estimation result corresponding to each feature element based on the bitstream, where the first probability estimation result includes a first peak probability; determining a set of first feature elements and a set of second feature elements from the plurality of feature elements based on a first threshold and the first peak probability corresponding to each feature element; and obtaining a decoded feature map based on the set of first feature elements and the set of second feature elements. This can improve encoding and decoding performance while reducing encoding and decoding complexity.