NBMP Envelope Descriptor for Redundant Task Deduplication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Moving Picture Experts Group (MPEG) Network Based Media Processing (NBMP) designs lack an application program interface (API) abstraction for network management and a common way to package messages, leading to inefficiencies in processing messages, especially in large-scale deployments where thousands or millions of messages may convey similar information, with order and source being crucial for effective processing.
Innovation Solution
The introduction of an envelope descriptor in the NBMP standard, which includes a message identifier, priority, sender ID, and other parameters, is added to all requests through the NBMP task API, enabling efficient processing by identifying and prioritizing repeated tasks and ensuring only one instance of each message is processed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If multiple messages are dispatched to convey task information in large-scale deployments, then complete information transmission is achieved, but processing efficiency deteriorates due to redundancy
Solution Approach 1:
The patent creates a standardized message template that serves as a reusable copy for multiple task dispatches. Instead of creating unique messages for each task, the system instantiates tasks from a single template with parameter substitution, eliminating redundant message creation while preserving complete information transmission.
Solution Approach 2:
The message template includes parameter placeholders that are dynamically substituted with specific task parameters. This allows the same message structure to convey different task information by changing only the parameter values, reducing message redundancy while maintaining information completeness.
2Productivity
If task information is packaged in standardized envelopes, then message processing efficiency improves, but API complexity increases
Solution Approach 1:
The envelope structure serves multiple functions simultaneously: it packages task information, identifies repeated tasks through templates, manages message routing, and handles prioritization. This multi-functionality reduces the need for separate mechanisms for each function, making the added structure worthwhile despite increased API complexity.
Solution Approach 2:
The message structure is segmented into distinct components: envelope metadata (template ID, parameters) and task-specific content. This segmentation allows the system to process different parts of the message independently, improving processing efficiency while organizing complexity into manageable segments.
3Reliability
If all task messages are processed individually, then message order and source are preserved, but processing time increases
Solution Approach 1:
The system performs preliminary actions by pre-defining message templates with all necessary information structure before actual task dispatch. This preliminary structuring allows for efficient matching and deduplication of repeated tasks, reducing processing time while maintaining reliability through the preserved template definitions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of processing media content in Moving Picture Experts Group (MPEG) Network Based Media Processing (NBMP), includes obtaining multiple tasks for processing the media content, each of the multiple tasks having an envelope descriptor indicating, for a respective one among the multiple tasks, a message identifier (ID), a message priority and a sender ID of a sender, determining whether repeated tasks exist among the obtained multiple tasks, based on the message ID and the sender ID of each of the obtained multiple tasks, based on the repeated tasks being determined to exist, processing the media content, using one among the repeated tasks that has a highest priority among the message priority of each of the repeated tasks, and based on the repeated tasks being determined to not exist, processing the media content, using the obtained multiple tasks.