Semantic Object Recognition Using Appearance-Behavior Mismatch

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems for object detection and identification in sensor networks often misclassify objects due to misclassification or malicious obfuscation, leading to inaccurate assessments of object types.

Innovation Solution

A device uses a semantic reasoning engine to determine both the appearance and behavioral type of an object based on sensor data, and provides an indication when the behavioral type does not match the appearance type.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated object detection and identification systems are used, then productivity is improved, but measurement precision deteriorates due to misclassification

Engineering Contradiction:

Improveautomated object detection efficiencyVSAvoidobject classification accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system implements feedback by comparing appearance-based classification results with behavior-based classification results. When discrepancies are detected between the two classification outcomes, the system generates alerts and can re-evaluate the appearance classification, creating a closed-loop verification mechanism that improves accuracy while maintaining automated processing.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system introduces behavior analysis as an intermediary verification layer between the appearance-based classification and the final object identification. This intermediary behavior assessment acts as a mediator that validates or challenges the appearance-based classification, reducing misclassification errors without eliminating automated processing.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Device complexity

If simple appearance-based classification is used, then device complexity is reduced, but reliability deteriorates due to misclassification and obfuscation

Engineering Contradiction:

Improveclassification system complexityVSAvoidobject identification accuracy

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The system merges appearance-based classification and behavior-based classification into a unified object identification process. By combining these two independent classification approaches, the system achieves more reliable object identification than either method could provide alone, while keeping each individual classification module relatively simple.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system creates a multi-functional classification framework where the same sensor data is processed through multiple classification pathways (appearance and behavior). This universal approach allows a single system to perform both appearance-based and behavior-based classification, improving reliability without requiring separate dedicated systems.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Measurement precision

If behavior analysis is added to appearance-based classification, then measurement precision is improved, but device complexity increases

Engineering Contradiction:

Improveobject classification accuracyVSAvoidclassification system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary behavior analysis using sensor data before finalizing object classification. By conducting behavior assessment in advance and comparing it with appearance-based classification, the system improves accuracy while organizing the complexity into a structured, multi-stage process that can be efficiently managed.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12518175B2Object recognition and behavioral analysis using semantic reasoning

Publication Date: 2026.01.06 CISCO TECHNOLOGY INC

AI summary

In one embodiment, a device detects an object in an area based on sensor data generated by a plurality of sensors deployed to that area. The device uses a semantic reasoning engine to determine an appearance type of the object, based on the sensor data. The device uses the semantic reasoning engine to determine a behavioral type of the object, based on the sensor data. The device makes, using the semantic reasoning engine, a determination that the behavioral type of the object does not match the appearance type of the object. The device provides an indication of the determination for display.