Modeling method for cross-inducing-mode emotion electroencephalogram recognition

Abstract

The invention discloses a cross-inducing-mode emotion electroencephalogram recognition and modeling method which comprises the following steps: designing inducing tests with different stimulation modes; collecting and preprocessing electroencephalogram data; extracting emotion electroencephalogram characteristics from the preprocessed electroencephalogram data; taking a part of video and picture inducing samples as a training set and the other part of the samples as a test set, and performing classification and recognition on emotion electroencephalogram characteristics corresponding to the training set by a support vector machine classifier based on recursive characteristic screening, wherein samples in the training set and the test set are from different material sources. According to the method, emotion electroencephalogram samples of different inducing modes are all tested, thus better conforming to practical application scenarios; the objective and moderate emotion recognition method can provide technical support for emotion detection on on-orbit astronauts, or enables a computer in a man-machine interaction field to have capabilities of sensing, recognizing and understanding human emotions so as to make a sensitive and friendly response to the emotion of a user and build a really intelligent and humanized man-machine environment.

Description

technical field [0001] The invention relates to the field of human-computer emotion interaction and the field of emotion monitoring for astronauts in the field of on-orbit spaceflight, in particular to a cross-induced mode emotion EEG recognition modeling method. Background technique [0002] Emotion is a high-level function of the human brain, which ensures the survival and adaptation of organisms, and has an important impact on individual learning, memory, and decision-making. Emotions are also a source of individual variation and are a key component of many personality traits and psychopathologies. With the development of society, people of all ages and fields are experiencing more and more emotional distress, and the correct identification and adjustment of emotions is of great significance to improving the quality of human life and ensuring physical and mental health. [0003] At present, researchers at home and abroad use facial expressions, voice, and posture signals...

AI Technical Summary

Problems solved by technology

However, there is a big limitation in this approach: the classification model established under a single emotion-evoked mode is likely to take into account some features that are only related to the stimulus attribute response and not related to emotion (that is, EEG features may only be related to the evoked mode, and does not contain emotion-specific information), the classification model cannot guarantee a good recognition ability for emotional EEG samples evoked by other stimulation patterns

Images