EEG Brainprint Recognition via Feature Disentanglement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing brainprint recognition methods based on EEG signals are task-dependent and have limitations, requiring specific physiological conditions and being less applicable in real-world scenarios due to their reliance on external stimuli, which restricts their robustness and usability.
Innovation Solution
A task-independent brainprint recognition method using feature disentanglement by decorrelation, which preprocesses EEG data to extract multi-scale time-frequency-space features and employs a primary brainprint and task disentangling neural network model to separate identity and task information, utilizing adversarial self-supervision for robust recognition across tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If task-dependent brainprint recognition methods are used, then recognition accuracy for specific tasks can be improved, but adaptability across different tasks deteriorates
Solution Approach 1:
The patent segments the EEG feature space into task-related components and identity-related components through decorrelation transformation. By separating these two types of information, the method can focus on identity features for recognition while eliminating task-dependent variations, thus achieving both high accuracy and cross-task adaptability
Solution Approach 2:
The patent transforms the feature representation by applying decorrelation to change the parameter space. This transformation modifies how features are organized and related, converting task-dependent features into task-independent identity features, thereby enabling the system to maintain high recognition accuracy across different tasks
2Reliability
If external stimulus-based brainprint recognition is used, then recognition can be achieved under controlled conditions, but ease of operation and applicability in real-world scenarios deteriorates
Solution Approach 1:
The patent extracts and removes the task-related information from the EEG features through decorrelation, isolating only the identity-related components. This extraction process eliminates the need for specific external stimuli or controlled task performance, making the system easier to operate in real-world scenarios while maintaining reliability
3Measurement precision
If task-specific brainprint recognition methods are used, then recognition performance for that task can be optimized, but device complexity and difficulty of deployment increase
Solution Approach 1:
The patent creates a universal recognition framework that can handle multiple tasks through a single decorrelated feature space. By designing the system to extract task-independent identity features, one deployment can serve multiple purposes across different tasks, reducing overall system complexity and deployment difficulty while maintaining optimized recognition performance
Data Source
AI summary
The present disclosure provides a task-independent brainprint recognition method based on feature disentanglement by decorrelation. Existing methods fail to mine inherent identity information of a brain, leading to poor robustness of brainprint recognition in a scenario across tasks and difficulty of promotion thereof in practical use. The present disclosure firstly uses two branch networks to perform coarse-grained decomposition of identity information and task related information in an electroencephalogram (EEG). Secondly, in consideration of an influence of a task state on the identity information, a decorrelating method is employed such that the identity information and the task related information are independent as much as possible. Finally, a brainprint feature in the EEG is fully utilized for classification by adversarial self-supervision. The method of the present disclosure is good in performance and capable of realizing efficient task-independent brainprint recognition, and is a brainprint recognition method robustly useful in the real life.