Spatial Model Assignment for Multi-Unit Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI systems struggle to efficiently utilize multiple computing units for object recognition tasks, leading to suboptimal performance and increased power consumption due to the lack of effective assignment of recognition models based on space and computing unit characteristics.
Innovation Solution
An electronic device is equipped with a plurality of computing units and recognition models, which divides a space into subset spaces based on object information, assigns specific recognition models to each subset space based on their characteristics, and allocates these models to computing units with matching capabilities, optimizing the use of resources for efficient object recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a single recognition model is used for the entire space, then the system is simple to operate, but the object recognition efficiency and accuracy are reduced due to inability to adapt to different space characteristics
Solution Approach 1:
The patent divides the entire space into multiple subset spaces based on spatial structure information and object distribution characteristics. Each subset space is then assigned a dedicated recognition model tailored to its specific characteristics, enabling optimized object recognition performance in each region while maintaining overall system manageability through automated segmentation processes.
Solution Approach 2:
The patent implements local quality by assigning different recognition models to different subset spaces based on their specific characteristics such as spatial structure, object density, and environmental features. This allows each region to receive the most appropriate model for its specific requirements, improving overall recognition efficiency while adapting to local variations in the environment.
2Productivity
If multiple recognition models are assigned to different computing units, then the system can optimize recognition performance for different spaces, but the device complexity increases due to model management and assignment overhead
Solution Approach 1:
The patent performs preliminary actions by pre-dividing the space into subset spaces and pre-assigning appropriate recognition models to computing units based on their characteristics before actual object recognition tasks begin. This preparation phase establishes a structured framework that simplifies runtime operations and reduces the complexity of dynamic model management during execution.
Solution Approach 2:
The system implements self-service through automated model selection and assignment processes that evaluate space characteristics and computing unit capabilities independently. The system automatically determines optimal model-computing unit mappings without requiring manual intervention, reducing operational complexity while maintaining high recognition accuracy across different spaces.
3Use of energy by moving object
If recognition models are assigned based on computing unit characteristics, then resource utilization efficiency is improved, but the system requires more sophisticated assignment algorithms increasing implementation complexity
Solution Approach 1:
The patent applies parameter changes by evaluating multiple characteristics of computing units (such as computational capability, power consumption, memory resources) and using these parameters to determine optimal model assignments. By dynamically adjusting model selection based on real-time monitoring of computing unit performance and energy consumption, the system optimizes power usage while managing implementation complexity through structured evaluation criteria.
Data Source
AI summary
An electronic device recognizing an object is provided. The electronic device includes a plurality of computing units, a memory, and a processor configured to control at least one of the plurality of computing units such that object information about objects obtained by recognizing the objects existing in a space by using a first recognition model, divide the space into a plurality of subset spaces, based on the object information, determine at least one recognition model, based on characteristic information of each of the subset spaces, assign the determined recognition model to one computing unit, based on characteristic information of each of a plurality of computing units and characteristic information of the determined recognition model, and control the plurality of computing units to perform object recognition by using the determined recognition model and the one computing unit in each of the subset spaces.