AI Object Detection With Retrieval Confirmation for Open Categories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection models struggle with accuracy when applied to large-scale image databases outside their training data sets, particularly due to the scale variance problem of objects and the challenge of handling open data categories.
Innovation Solution
A dual-stream detection approach combining an object detection model and an object retrieval model, where the detection model performs initial feature extraction and location prediction, followed by the retrieval model for enhanced category confirmation, using a feature pyramid network for multi-scale feature fusion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single object detection model is used for feature extraction and category prediction, then the detection speed is maintained, but the accuracy deteriorates when applied to large-scale image databases outside training data sets
Solution Approach 1:
The detection system is segmented into two independent streams: an object detection model for initial feature extraction and location prediction, and an object retrieval model for enhanced category confirmation. This segmentation allows each model to specialize in specific tasks, improving overall accuracy while maintaining manageable complexity through modular architecture.
Solution Approach 2:
The patent combines the outputs of two separate models (detection model and retrieval model) through confidence level fusion. The detection model provides first confidence levels for initial predictions, while the retrieval model provides second confidence levels for verification, and their combined results achieve higher accuracy than either model alone.
2Adaptability or versatility
If multi-scale feature extraction is performed to handle objects of different sizes, then the detection coverage is improved, but the computational complexity increases
Solution Approach 1:
The feature pyramid network performs multi-scale feature extraction in advance, generating feature maps at multiple scales before the actual detection and retrieval processes. This preliminary multi-scale feature preparation enables the system to handle objects of various sizes efficiently without repeating computationally expensive operations during inference.
3Measurement precision
If a dual-model approach is used with confidence level fusion, then the category prediction accuracy is improved, but the detection time increases
Solution Approach 1:
The system performs partial verification by the retrieval model based on the detection model's initial predictions. Rather than requiring both models to process all cases equally, the retrieval model focuses on confirming categories for detected objects, performing excessive verification only where needed to improve accuracy without proportionally increasing processing time for all cases.
Data Source
AI summary
This application discloses an artificial intelligence-based object detection method and apparatus. The method includes inputting a target image comprising an object to an object detection model; obtaining feature images of different scales from the target image using the object detection model; determining image location information of the object and a first confidence level that the object belongs to each category; acquiring a target region in which the object is located; inputting the target region to an object retrieval model, comparing the target region with sample images of a plurality of categories to obtain a second confidence level that the object belongs to each category; and determining a target category of the object based on the first confidence level and the second confidence level, a sum of the first confidence level and the second confidence level of the target category being the largest of the plurality of categories.