Object Recognition Model Selection for Cloud Audits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer vision systems require multiple object recognition models to identify various objects within digital images and videos, leading to inefficiencies and computational delays, especially when multiple compliance audits are performed simultaneously.
Innovation Solution
A cloud-based object recognition service that efficiently selects a minimum set of object recognition models by using a non-iterative method to determine the necessary models for object recognition, reducing computational complexity and improving audit result generation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple object recognition models are used to identify various objects within digital images and videos, then the comprehensiveness of object recognition is improved, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the object recognition task by dividing it into multiple specialized models, each trained to recognize specific object categories. This segmentation allows the system to achieve comprehensive object recognition while managing computational complexity by processing only relevant models for each image, rather than running all models on every image.
Solution Approach 2:
The patent implements partial action by selecting and applying only the necessary subset of object recognition models based on the specific requirements of each compliance audit task. This avoids the excessive computation of running all available models regardless of need, thereby reducing processing time while maintaining recognition comprehensiveness.
2Adaptability or versatility
If multiple object recognition models are deployed for comprehensive object identification, then the coverage of recognizable objects is improved, but the processing speed decreases
Solution Approach 1:
The patent performs preliminary actions by pre-training multiple specialized object recognition models for different object categories before deployment. During compliance audits, the system quickly selects the appropriate pre-trained models based on the audit requirements, avoiding the need to train models on-demand and significantly improving processing speed while maintaining comprehensive object coverage.
3Measurement precision
If a comprehensive set of object recognition models is used, then the accuracy of object identification is improved, but the computational resources required increase
Solution Approach 1:
The patent applies local quality by training different object recognition models with specialized architectures and parameters optimized for specific object categories. Each model has local quality tailored to its designated object type, achieving high accuracy for that category while consuming fewer computational resources than a general-purpose model would require for the same task.
Data Source
AI summary
Disclosed herein are system, method, and computer program product embodiments for compliance auditing using cloud based computer vision. In one aspect, a system is configured to receive, from a mobile device, a compliance audit request to at least recognize one or more products within the audit image. The system is further configured to select a first object recognition model having a first associated object recognition model identifier from a model selection list based at least on a required object recognition list, wherein the first object recognition model is configured to recognize a first set of object names within the required object recognition list. The system is further configured to request the computer vision system to perform object recognition using the first object recognition model to recognize the first set of object names within the audit image, and transmit audit result information to the mobile device.