Similarity-Based Object Grouping for Consistent Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In object recognition scenarios, similar objects are often recognized by different entities leading to inconsistent recognition results and low efficiency due to random allocation, which affects accuracy and consistency.
Innovation Solution
Objects are grouped based on similarity and allocated to a target entity that matches the group for recognition, using similarity metrics and trained models to determine the optimal entity for processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If objects are allocated to different entities for recognition, then recognition tasks can be distributed and processed in parallel, but similar objects may be recognized by different entities leading to inconsistent recognition results
Solution Approach 1:
The patent groups similar objects together into the same recognition task and assigns them to a single entity. This merging approach ensures that all similar objects are processed by the same entity using consistent recognition criteria, thereby maintaining consistency of recognition results while still enabling parallel processing of different object groups across multiple entities.
2Productivity
If similar objects are processed by different entities, then workload can be distributed across multiple entities, but recognition accuracy decreases due to inconsistent results
Solution Approach 1:
The patent combines similar objects into unified recognition tasks that are assigned to single entities. This ensures that recognition accuracy is maintained through consistent processing of similar objects, while the overall system processing capacity is preserved by distributing different object groups to multiple entities in parallel.
3Device complexity
If random allocation is used to distribute objects to entities, then task distribution is simple and fast, but recognition efficiency is low due to inconsistent handling of similar objects
Solution Approach 1:
The patent performs preliminary grouping of similar objects before allocation to entities. By pre-processing the object list to identify and group similar objects together, the system enables more efficient allocation decisions that improve recognition efficiency, while the grouping mechanism itself remains relatively simple to implement.
Data Source
AI summary
According to embodiments of the disclosure, a method, an apparatus, a device, and a storage medium for object recognition are provided. The method includes: obtaining an aggregation result of a plurality of objects, the aggregation result including at least one group of objects aggregated based on a similarity; determining a target entity that matches the at least one group of objects for performing object recognition; and providing the at least one group of objects to the target entity. In this way, similar objects can be provided to a matched entity for recognition, thereby improving recognition efficiency and improving accuracy and consistency of recognition results.