Object Recognition Using Reference Probabilities for Low-Confidence Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition systems face challenges in achieving high accuracy due to the complexity of monitoring scenes and varying object statuses, leading to low recognition performance.
Innovation Solution
A system that adjusts initial recognition values based on reference probabilities derived from a reference image set, improving the accuracy of object categorization by refining the initial recognition results when certain conditions are met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a recognition model is used to determine object categories based on images, then the system can automatically perform object recognition, but the recognition accuracy is low due to scene complexity and object status variations
Solution Approach 1:
The patent applies feedback by using reference probabilities derived from reference images to correct and refine the initial recognition results. The system compares the initial recognition values against reference probabilities and adjusts the results accordingly, creating a feedback loop that improves accuracy while maintaining automation.
Solution Approach 2:
The patent uses preliminary action by pre-calculating reference probabilities from a reference image set before actual object recognition. These reference probabilities are stored and ready to be applied when processing new images, allowing the system to improve recognition accuracy without adding significant computational delay during the actual recognition task.
2Measurement precision
If the initial recognition values are adjusted using reference probabilities, then the recognition accuracy is improved, but the system complexity increases due to additional processing steps
Solution Approach 1:
The reference probabilities are calculated in advance from a reference image set and stored for later use. This preliminary computation allows the actual recognition process to simply involve comparing initial recognition values against pre-computed references, significantly reducing the complexity of the real-time processing steps.
Solution Approach 2:
The system uses reference images and their corresponding reference probabilities as copies or templates that represent typical object categories. By comparing actual object images against these reference copies, the system can improve accuracy without requiring complex analysis of every pixel and feature in detail.
Data Source
AI summary
The present disclosure relates to systems and methods for object recognition. The systems may obtain an image associated with an object. The systems may determine, using a recognition model, an initial recognition result of the object based on the image. The initial recognition result may include a plurality of initial recognition values corresponding to a plurality of predetermined categories respectively. Each of the plurality of initial recognition values may indicate an initial probability that the object corresponds to a corresponding predetermined category. In response to determining that the initial recognition result satisfies a preset condition, for each of the plurality of predetermined categories, the systems may obtain a reference probability corresponding to the predetermined category. The systems may determine a target recognition result of the object by adjusting the plurality of initial recognition values based on a plurality of reference probabilities corresponding to the plurality of predetermined categories respectively. The target recognition result may include a plurality of target recognition values corresponding to the plurality of predetermined categories respectively. Each of the plurality of target recognition values may indicate a target probability that the object corresponds to a corresponding predetermined category.