Compact Function Descriptors for Multi-Instance NBMP Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Network-based Media Processing (NBMP) standard faces inefficiencies due to large Function descriptors for popular Functions, which can lead to slower retrieval and limited flexibility in media processing, as the size of Function descriptors grows linearly with the number of Function Groups they belong to, restricting additional usage parameters.
Innovation Solution
Implementing a method for multi-instance NBMP function execution that allows for compact Function descriptors, enabling different instances of a Function to have unique restrictions, independent of the number of Function Groups, and supporting additional restrictions beyond input/output/configurations, thereby improving retrieval efficiency and flexibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Function descriptors include all restrictions for every_function Group they belong to, then completeness of function information is improved, but descriptor size grows linearly with the number of_function Groups
Solution Approach 1:
The patent segments the function descriptor into two parts: a compact base descriptor containing common restrictions applicable to all function groups, and separate function group-specific restriction sets. This segmentation allows the base descriptor to remain small while completeness is maintained through the modular addition of group-specific restrictions only when needed.
Solution Approach 2:
The patent applies local quality by making restrictions local to each function group context rather than globally duplicated. Each function group can specify only the restrictions that are relevant to its particular usage, while inheriting common restrictions from the base descriptor. This reduces overall descriptor size while maintaining local completeness.
2Reliability
If Function descriptors grow with the number of_function Groups, then all restrictions are captured, but retrieval speed decreases
Solution Approach 1:
By segmenting the descriptor into a compact base portion and optional group-specific extensions, the system enables rapid retrieval of the common base descriptor while allowing selective loading of group-specific restrictions only when required for that particular function group context.
Solution Approach 2:
The base descriptor with common restrictions is prepared and stored in advance in a compact form. When a function is retrieved, the base descriptor is immediately available, and group-specific restrictions are added only if needed, rather than retrieving and processing all possible restrictions for every function group.
3Adaptability or versatility
If multiple restrictions are added for popular Functions, then flexibility in media processing is improved, but descriptor complexity increases
Solution Approach 1:
The patent enables different restrictions to be applied locally to different function group contexts for the same function. Popular functions can have multiple sets of group-specific restrictions, each tailored to particular media processing scenarios, while the base descriptor remains simple and unchanged.
Solution Approach 2:
The base descriptor serves as a universal template that can be applied across multiple function groups. Popular functions can be reused with different group-specific restriction sets, allowing the same function to serve multiple purposes with different restrictions without increasing the base descriptor complexity.
Data Source
AI summary
A method, computer program, and computer system is provided for multi-instance Network-Based Media Processing (NBMP) function execution. A function call corresponding to an NBMP request is received. A function is retrieved from among a function group having a function descriptor, the retrieved function corresponding to the received function call. One or more instances of the retrieved function are executed.