Visual Scene Feature Extraction via Semantic Graph Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computing infrastructures face challenges in interpreting and making sense of the vast amount of visual scene data collected from multiple sources, including camera systems and sensors, which complicates the identification and classification of objects within these scenes.
Innovation Solution
The integration of object identification with semantic scene graphs that associate contextual data to feature spaces, allowing for the estimation of semantic behaviors by linking objects and their attributes through multi-feature graphs and association edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple camera systems and sensors are used to collect visual scene data, then the quantity and quality of observed data is improved, but the complexity of interpreting and processing this data increases
Solution Approach 1:
The patent segments the complex visual scene data into distinct feature spaces (e.g., object features, scene features, temporal features) and processes them independently through separate computational graphs. This segmentation allows the system to handle large quantities of data from multiple sensors by dividing the processing task into manageable, parallelizable components, thereby reducing the overall complexity of data interpretation while maintaining comprehensive data utilization
2Measurement precision
If object identification is integrated with semantic scene graphs, then the accuracy of behavior estimation is improved, but the computational resources and processing time required increase
Solution Approach 1:
The patent implements preliminary action by pre-building and maintaining hierarchical computational graphs that encode semantic relationships between objects, scenes, and behaviors before actual analysis occurs. These pre-organized knowledge structures enable rapid querying and inference during real-time processing, significantly reducing the computational time required for behavior estimation while maintaining high accuracy through the pre-established semantic frameworks
Solution Approach 2:
The system dynamically adjusts the level of detail and computational depth based on the specific analysis requirements and available resources. The computational graphs can be pruned, expanded, or reweighted in real-time to balance between processing speed and estimation accuracy, allowing the system to adapt to varying operational conditions without compromising the core accuracy-performance relationship
Data Source
AI summary
A computer system executing code instructions of a visual scene feature extractor and classifier system comprising a processor executing object recognition on a captured image of a visual scene of interest including a plurality of objects in the visual scene of interest to identify a first object from the plurality of objects with a set of feature attributes wherein the feature attributes include an image view of the first object where the processor executing code instructions for the visual scene feature extractor and classifier to generate a first feature vector value for a first feature attribute of the first object. The processor to generate a first feature space including a neighborhood of labeled, previously-observed first feature vector values around the first feature vector value and to generate a second feature vector value for a second feature attribute of the first object, wherein the second feature vector value is within a second feature space including a neighborhood of labeled, previously-observed second feature vector values and the processor to generate association edges between the first and second feature vector values and the first object and between labeled, previously-observed first and second feature vector values and candidate objects corresponding to previous observations to establish a multiple feature graph for the first object for determining weighted likelihood that the first object may be identified as at least one previously observed candidate object.