Human Emotional State Classification based on Wavelet Analysis and Statistical Feature Selection
Due to a major shortage of nurses in the U.S., it is possible that future healthcare service robots will be required to interact directly with patients. With this in mind, there is a need to design nursing robots with the capability to detect and respond to patient physical and emotional states and to facilitate positive patient experiences in healthcare. The objective of this research was to develop a new computational algorithm for human emotional state classification to facilitate effective patient-robot interaction (PRI). A simulated medicine delivery experiment was conducted in nursing homes using a service robot with different human-like features. Physiological signals, including heart rate (HR) and galvanic skin response (GSR), and subjective ratings of valence (happy-unhappy) and arousal (excitedbored) were collected on elderly residents during the experiment trials. These two types of data were used in a three-stage emotional state classification algorithm, including: (1) physiological feature extraction; (2) statistical-based feature selection; and (3) a machine learning model of emotional states. A pre-processed HR response was used; however, GSR signals were non-stationary and noisy and were further processed using wavelet analysis. A set of wavelet coefficients was used to represent an entire signal and to reveal time, amplitude and frequency of GSR features. A stepwise regression approach was used to select significant wavelet features, relating to each class of subjective emotional state, before being input into the machine learning model. Results of the statistical and neural network-based analyses indicated strong non-linear relationships between the physiological variables and emotional states. Arousal was significantly explained by HR and GSR while valence was explained by GSR alone. The new algorithm may serve as an effective method for service robot real-time detection of patient emotional states that could be used as a basis for adaptive behaviors to promote positive patient healthcare experiences.
Manida Swangnetr David Kaber Yuan-Shin Lee
Edward P. Fitts Department of Industrial & Systems Engineering, North Carolina State University,Rale Edward P. Fitts Department of Industrial & Systems Engineering, North Carolina State University, Ral
国际会议
17th World Congress on Ergonomics(第十七届国际人类工效学大会)
北京
英文
1-8
2009-08-09（万方平台首次上网日期，不代表论文的发表时间）