Design And Analysis Of Emotion Recognition Experiment

The World Health Organization points out that health is a state of perfect physical,mental and social adaptation.With the improvement of people’s living standard,they have a higher pursuit for spiritual enrichment and health.At present,the social competition is fierce,the pressure is too large,people’s mental health problems are more and more prominent,and the individual’s recent emotional state is an important indicator to observe their mental health status.Based on our current social perception,when people encounter psychological problems or long-term depression,they often lack relevant recourse,are ashamed to see a psychologist,or are too busy to pay attention to them.In this context,this paper proposes to use the wristband to collect physiological signals to identify people’s emotions,so that people can timely understand their emotional health.There have been international and domestic researches on emotion recognition using physiological signals,but professional medical equipment is mostly used for data collection.In order to realize the data collection of physiological signals on the bracelet,this subject selects skin electrical signals and heart rate signals for research.In the research of this subject,the relevant physiological signals of the subjects were collected by designing the emotional excitation scene and process.In this study,three different emotional states,neutral,positive and negative,were used to stimulate emotion.A total of three experimental data were collected,and 20 samples were collected in each experiment.Two groups of comparative experiments were carried out,in the material comparison experiment,the same group of subjects use different stimulating materials for emotional stimulation,while in the personnel comparison experiment,different subjects use the same stimulating materials for emotional stimulation.After data preprocessing and feature extraction,different algorithm models support vector machine,random forest and Ada Boost are used to analyze the data.Through the comparative analysis of these results,the relevant experimental conclusions are obtained.The results show that the recognition accuracy of strong emotional stimulation is significantly higher than that of weak emotional stimulation.When the emotion excitation intensity is strong,the recognition accuracy of skin electrical signal is better than that of heart rate signal.When the emotion excitation intensity is weak,the recognition accuracy of heart rate signal is significantly better than that of skin electrical signal.