Development of Multi-modal Physiological Signals Acquisition Platform and Its Application in Emotional Recognition Based on a Small Number of Samples

Emotion is a typical psychological process with obvious physiological characteristics. Physiological signals are difficult to imitate and can reflect real emotions, so the research of emotion recognition based on physiological signals is necessary. Multi-modal physiological signals are complementary, which have better performance than single-modal signals in emotion recognition. Because traditional multi-modal physiological signals acquisition device is complex or not suitable for portable products development, a multi-modal physiological signals acquisition platform with simple structure is designed in this paper to collect heart rate, ECG and body temperature. 52 features are extracted from multi-modal physiological signals in time domain and Mutual Information Feature Selection (MIFS) method is used in feature selection to obtain subset with better recognizability. Support Vector Machine (SVM) model for classifying excited state in game scene and calm state in non-game scene is trained and tested based on 60 samples. The top emotion recognition rate is 80% when 19 features are selected. The experimental results show that the multi-modal physiological signals acquisition platform built in this paper is effective and the emotions of subject can be recognized in specific scenes based on a small number of samples.