Continuous Learning Anomaly Detection Model Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing anomaly detection and state classification methods fail to adapt effectively to environmental changes, leading to misclassification of normal data as anomalies when the environment shifts.
Innovation Solution
A method and apparatus for continuous learning that involves generating feature vector matrices from anomaly detection data, calculating restoration errors, and updating detection and classification networks using stored normal and category data to adapt to environmental changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a typical anomaly detection method learns data of a normal category for a specific environment and performs anomaly detection after learning once, then the method properly performs anomaly detection in stable environments, but if the environment changes, the method causes errors in misjudging normal category data of the changed environment as anomalies
Solution Approach 1:
The patent implements continuous learning by dynamically updating the neural network model with newly acquired normal data over time. The model transitions from a static, one-time learning approach to a dynamic system that continuously adapts its parameters through iterative training on accumulating normal data, enabling it to adjust to environmental changes while maintaining reliable anomaly detection performance
Solution Approach 2:
The patent establishes continuous learning as an ongoing process where the system continuously acquires normal data, retrains the model, and updates its detection capabilities. This continuous cycle of learning and updating ensures the model remains adapted to current environmental conditions, resolving the contradiction between maintaining detection reliability and adapting to environmental changes
Data Source
AI summary
According to the present invention, a method for continuous learning of object anomaly detection and state classification model includes acquiring, by a detection and classification apparatus, information about a medium of anomaly detection from an inspection target; generating, by the detection and classification apparatus, an input value, which is a feature vector matrix including a plurality of feature vectors, from the medium information; deriving, by the detection and classification apparatus, a restored value imitating the input value through a detection network learned to generates the restored value for the input value; determining, by the detection and classification apparatus, whether a restoration error indicating a difference between the input value and the restored value is greater than or equal to a previously calculated reference value; and storing, by the detection and classification apparatus, the input value as normal data upon determining that the restoration error is less than the reference value.