Heterogeneous label space transfer learning method for brain-computer interface calibration

Abstract

The invention discloses a heterogeneous label space transfer learning method for brain-computer interface calibration, and relates to the field of brain-computer interfaces. The method comprises the following steps: marking and grouping an electroencephalogram signal sample set of a new user, and calculating an average covariance matrix in each group; grouping all samples of the electroencephalogram signal sample set of the auxiliary user according to label categories, and calculating an average covariance matrix in each group; according to the average covariance matrix meeting the set corresponding relation, transforming samples of the auxiliary users, assigning labels of the new users to the samples of the auxiliary users according to the corresponding relation, and obtaining transformedauxiliary user data; and combining the transformed auxiliary user data and the marked new user samples as a training set, and constructing a machine learning model on the training set. According to the method, the model learning capability of the new user can be improved by means of the auxiliary user data of the heterogeneous label space, and the calibration data required by the new user is greatly reduced, so that the calibration time is effectively reduced.

Description

technical field [0001] The invention belongs to the field of brain-computer interface, and more specifically, relates to a method for learning transfer of heterogeneous label spaces for brain-computer interface calibration. Background technique [0002] Brain-computer interface is a system that provides a direct interaction channel between the brain and external devices (such as computers, robots, etc.), and is regarded as the ultimate form of human-computer interaction. It not only allows users to directly control the movement of external devices through brain signals, such as mechanical exoskeletons or drones; it can also be used to judge the current state of the brain, such as sleep status, fatigue and so on. The input signals of the brain-computer interface system include EEG, magnetoencephalogram, functional MRI, etc., among which EEG is the most common. EEG is usually collected from the surface of the scalp by a head-mounted EEG cap through electrodes, and the dischar...

AI Technical Summary

Problems solved by technology

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a heterogeneous label space transfer learning method for brain-computer interface calibration. When the user's label space is different, the auxiliary user data cannot be used, and only a large amount of new user data can be collected for calibration, resulting in a technical problem that the calibration time is greatly prolonged

Images