Analysis Apparatus Discriminating Feature Classes Using Multiple Discriminators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing discrimination methods for action types in sensing data face challenges due to irrelevant information, such as background data and shared features, leading to degraded accuracy and difficulties in determining transition timing between actions, which limits the improvement of discrimination performance despite increased quality of learning data.
Innovation Solution
An analysis apparatus that uses multiple discriminators to identify specific features by determining data portions where one discriminator's feature is present but others are not, and where all discriminators agree on shared features, improving the accuracy of feature class discrimination and generating high-quality learning data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple discriminators are used to improve discrimination accuracy, then measurement precision improves, but device complexity increases
Solution Approach 1:
The system segments the discrimination task by deploying multiple specialized discriminators, where each discriminator is trained to recognize specific action types. This segmentation allows each discriminator to focus on particular features relevant to its designated action type, improving overall discrimination accuracy while maintaining manageable complexity through modular design
Solution Approach 2:
The system merges the outputs of multiple discriminators through a determination unit that integrates their results. By combining the discrimination outcomes from multiple specialized discriminators, the system achieves higher measurement precision through collective decision-making, resolving the contradiction between using multiple components and maintaining system simplicity
2Reliability
If learning data quality is improved by capturing actions under various conditions, then discrimination performance improves, but loss of time increases due to difficult transition timing determination
Solution Approach 1:
The determination unit provides feedback by analyzing the outputs of multiple discriminators and identifying transitions when discriminator agreements change. This feedback mechanism automatically determines transition timing based on the collective behavior of discriminators, eliminating the need for manual timing determination and reducing time loss while maintaining reliable discrimination performance
3Adaptability or versatility
If learning data includes information from transition periods between actions, then adaptability improves, but measurement precision deteriorates due to mixed action information
Solution Approach 1:
The system applies partial action by having each discriminator focus on recognizing only its specific action type rather than attempting to identify all action types simultaneously. During transition periods, this partial recognition approach allows the system to capture adaptability information while maintaining precision, as each discriminator reports only on its specialized detection without being confused by other action types
Data Source
AI summary
An analysis apparatus according to one or more embodiments may identify the classes of features included in object data using a plurality of discriminators that are respectively configured to discriminate the presence of features of classes different to each other; and determines that a first data portion, with respect to which discrimination is established by one of the plurality of discriminators, but discrimination is not established by the remaining discriminators, includes a feature of the particular class that is discriminated by the one discriminator, and determines that a second data portion, with respect to which discrimination is established by all of the discriminators including the one discriminator, does not include a feature of that particular class.