Predictive Configuration Records for Low-Overhead Video Codecs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media coding systems inefficiently handle parameter sets, leading to significant overhead due to repetitive signaling of parameter information, especially in band-limited environments, which affects coding efficiency and quality.
Innovation Solution
Implementing configuration records that can be developed predictively from other configuration records, with independent and dependent types, and using memory management techniques to store and manage these records, allowing for bandwidth savings by only updating changed parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If parameter sets are signaled frequently to maintain up-to-date coding information, then coding accuracy is improved, but bandwidth overhead increases significantly
Solution Approach 1:
The parameter set is divided into a template portion (containing common parameters) and a non-template portion (containing specific parameters). The template is signaled once and reused, while only the non-template portion is updated frequently, reducing bandwidth overhead while maintaining coding accuracy.
Solution Approach 2:
A template parameter set is established in advance that contains common parameters applicable to multiple coding units. This template is signaled once and then reused for subsequent coding units, eliminating the need to re-signal these common parameters and reducing bandwidth overhead.
2Reliability
If complete parameter sets are transmitted for each coding unit to ensure decoding accuracy, then decoding reliability is improved, but bandwidth consumption increases
Solution Approach 1:
The parameter set structure is organized with a template portion containing common parameters that nest across multiple coding units, and a non-template portion containing unit-specific parameters. This nested structure allows efficient inheritance and reuse, ensuring decoding reliability while reducing bandwidth consumption.
Solution Approach 2:
The template parameter set is copied and reused across multiple coding units. Instead of transmitting complete parameter sets for each unit, the system copies the template and only transmits modifications in the non-template portion, ensuring decoding reliability while minimizing bandwidth consumption.
3Ease of operation
If parameter information is repeated in every coding unit to ensure self-contained decoding, then decoding independence is improved, but bitstream overhead increases
Solution Approach 1:
The parameter information is segmented into template parameters (common to multiple units) and non-template parameters (specific to each unit). The template is signaled once to ensure decoding independence, while only the non-template portion is repeated in each coding unit, reducing bitstream overhead.
Solution Approach 2:
The template parameter set is established in advance before the coding units. This preliminary action ensures that decoders can independently reconstruct the common parameters without needing them repeated in each unit, while only unit-specific parameters are included in each coding unit's data.
Data Source
AI summary
The present disclosure describes improvement in the design of syntaxes that carry bandwidth-compressed media between encoders and decoders. In particular, it relates to improvements in representation of parameter sets and a prediction process between parameters sets. According to embodiments of the disclosure, a coding system represents data or other information according to a syntax that is categorized and organized into types and hierarchies, where information of the lower layers or levels of the hierarchy are predicted from information of the higher levels. Although the present discussion discusses primarily video applications, the principles of the present disclosure find application with other types of data that may be partitioned into units, groups of units, or layers, and for which parameter set information may be present at different stages/levels of the information unit hierarchy. For example, such data could include audio data, mesh or point-cloud information, text, among others.