Method to Identify Emotions in Immersive Virtual Learning Environments Using Head and Hands Spatial Behavioral Information

Emotion detection has been carried out using different channels and sensors, facial recognition being one of the most frequent. Likewise, physiological signals such as EEG, ECG, EMG, and GSR along with various artificial intelligence classification techniques, such as SVM and NN, are recurrent for this task. However, to achieve this, multiple devices are used, most are considered invasive, expensive, not very portable and require large computing capacity. Therefore, and supported by the literature that suggests that different cognitive states are reflected in behavior and body postures, this work proposes an emotional detection approach during the use of an Immersive Virtual Learning Environment (IVLE), using the information provided by the Immersive Virtual Reality (IVR) device in conjunction with data obtained during interactions and adapted self-report instruments. Since only the data extracted from the device itself is used, we consider this as a single-modal approach. This seeks to be a portable-low-cost solution with precision enough so it can be implemented in an affective tutoring system in an IVR IVLE.