Multi-region Person Recognition Certainty Update
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional camera systems that capture images from multiple locations often produce erroneous person recognition due to the lack of a candidate with high recognition likelihood, leading to incorrect overall recognition results when spatial constraints are not strictly adhered to.
Innovation Solution
An information processing apparatus that estimates the certainty of a person's identity across multiple regions by updating person certainty factors based on temporal and spatial constraints, adjusting the certainty based on the duration a person is recognized in different areas, ensuring accurate recognition even without a conspicuously high recognition likelihood in a given environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If recognition results are fixed based on spatial constraint without considering recognition likelihood, then processing speed is improved, but recognition accuracy deteriorates
Solution Approach 1:
The patent changes the parameter of certainty factor from a fixed binary decision to a dynamic value that evolves over time based on recognition likelihood and temporal duration. The certainty factor is updated continuously as new recognition results arrive, allowing the system to transition from quick but potentially inaccurate fixed decisions to more accurate time-evolving decisions.
Solution Approach 2:
The patent introduces dynamics by making the recognition result flexible rather than fixed. The certainty factor dynamically adjusts based on the duration a person is recognized in a region and the accumulation of recognition evidence from multiple cameras, allowing the system to adapt between speed and accuracy based on the confidence level.
2Device complexity
If recognition candidates are narrowed down using spatial constraint, then device complexity is reduced, but recognition reliability deteriorates
Solution Approach 1:
The patent implements feedback by continuously updating the certainty factor based on new recognition results from multiple cameras. The system monitors whether a person is recognized in different regions over time and adjusts the certainty factor accordingly, providing feedback that improves reliability without significantly increasing complexity.
Solution Approach 2:
The patent performs preliminary action by establishing spatial constraints and initial certainty factors before final recognition decisions are made. This preliminary framework allows the system to quickly filter unlikely candidates while maintaining the ability to revise decisions as new evidence emerges, balancing complexity and reliability.
3Loss of time
If a single recognition result is fixed as correct, then processing time is reduced, but loss of information increases
Solution Approach 1:
The patent maintains continuity of useful action by continuously updating the certainty factor as new recognition information becomes available. Instead of stopping after a single fixed decision, the system continues to process recognition results from multiple cameras and time points, accumulating information without significant additional processing time.
Solution Approach 2:
The patent applies partial action by updating the certainty factor only when necessary based on the duration and consistency of recognition results. The system performs minimal updates when confidence is already high, avoiding excessive processing while still capturing important recognition information when it becomes available.
Data Source
AI summary
An information processing apparatus includes an estimation unit configured to estimate, based on a feature of a person of interest contained in an image captured in each of a plurality of regions and a feature of a previously set registered person, a certainty that the person of interest is the registered person, and an updating unit configured to, with respect to a first person of interest subjected to image capturing in a first region out of the plurality of regions, in a case where a period for which a state in which a certainty that a second person of interest contained in an image captured in a second region different from the first region is the registered person is larger than a threshold value is kept is longer than a predetermined period, perform updating in such a way as to lower a certainty that the first person of interest in the first region is the registered person, and, in a case where the period for which the state in which the certainty that the second person of interest is the registered person is larger than the threshold value is kept is shorter than the predetermined period, perform updating in such a way as not to lower the certainty that the first person of interest in the first region is the registered person.