Event Cache Management for Unauthorized Subject Alert Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video analytic technologies face challenges in generating consistent alerts for unauthorized subjects due to varying environmental and imaging conditions, leading to inaccurate and unnecessary alerts.
Innovation Solution
A method and system that manage events to generate alerts by storing data identifying a subject in a cache for a pre-determined period, detecting recall events indicating the subject is not likely to be unauthorized, and removing the event from the cache within that period to prevent false alerts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional video analytic technology generates alerts immediately upon detecting a subject that does not match authorized subjects, then the system responds quickly to potential threats, but the accuracy of alerts deteriorates due to false positives from varying environmental and imaging conditions
Solution Approach 1:
The system performs preliminary actions by storing detection events in a cache with timestamps before generating final alerts. When a subject is detected that does not match authorized subjects, the event is cached for a predetermined period rather than immediately triggering an alert. This preliminary caching allows for potential recall events to override false positives, thereby improving alert accuracy while maintaining relatively quick response times.
2Reliability
If the system caches events for a predetermined period to verify unauthorized status, then alert accuracy improves by reducing false positives, but the response time deteriorates due to the verification delay
Solution Approach 1:
The system applies partial action by caching events for a predetermined period rather than requiring full verification before alert generation. The cache duration is optimized to be sufficient to catch recall events for false positives (improving accuracy) but not so long as to significantly delay legitimate alerts (minimizing time loss). This balanced approach resolves the contradiction between accuracy and response time.
3Productivity
If the system immediately generates alerts for subjects that do not match authorized subjects, then potential threats are identified quickly, but unnecessary alerts increase due to inconsistent detections under varying conditions
Solution Approach 1:
The system implements feedback by monitoring for recall events within the predetermined cache period. When a recall event occurs (indicating the subject may have been misidentified), the system feedbacks this information to override the cached event and prevent false alert generation. This feedback mechanism reduces false alert rates while maintaining efficient threat detection by quickly processing and validating events.
Data Source
AI summary
Present disclosure provides a method for managing an event to generate an alert indicating a subject is likely to be unauthorized, the event comprising data identifying the subject, the method comprising: within a pre-determined period based on a timestamp at which the event is stored in a cache, detecting a recall event comprising the data identifying the subject, the recall event indicating the subject is not likely to be unauthorized; and removing the event from the cache in response to detecting the recall event comprising the data identifying the subject within the pre-determined period.