Options
2016
Journal Article
Titel
Neurocognitive tools for assessing haptic interaction
Titel Supplements
Abstract
Abstract
Haptic interaction is based on applying forces, vibrations, and/or motions to the computer user with special haptic devices. To understand how believable the haptic interaction is, questionnaires evaluating sensory immersion, comfort, realism, and satisfaction are used. However, this method only allows for overall evaluation of the quality of experience. Based on the existing ways of using EEG for identification of the brain states, we hypothesized that we should be able to capture changes of the users' feelings during haptic interaction while traditional user questionnaires can be used as a reference. Moreover, using time-stamped EEG data at any time interval starting from 1/32 sec, we should also be able to actively engage the potential users into the design and development phases of various haptic interactions. To prove the hypothesis, we proposed, implemented and applied neurocognitive methods, based on recognition from EEG the emotion and stress levels, in a project where haptic interaction was added to networked video conversations with Skype. In our study, 10 participants were grouped in pairs based on their cultural and educational similarities. After introduction to haptic technology followed by EEG calibration, each pair of the testers was asked to participate in common Skype conversation followed by 3 more Skype conversations where they also had to physically interact with each other (while sitting in different rooms) using haptic devices: they could see each other hands and physically feel their motions. These haptic Skype conversations were organized into 3 trials with equal as well as dominating roles of the participants. EEG data was recorded for each of the 4 trials with Emotiv Epoc + device. The Stroop test was used to label EEG data. The questionnaire to assess the subject emotions, mental workload and stress was also given. The EEG signals of all the participants were processed to recognize emotional states and stress levels by two ways: as overall averaged levels over each trial, as well as averaged by 10 sec values for detailed and quantified measurements. After analyzing the results, we have concluded that the EEG-based emotion and stress recognition results match the results collected with traditional questionnaires; however, the EEG-based results also allow us to do data analyses during the task performance. Thus, in the EEG-data we could observe the specifics of the behaviors of experienced haptic users, as well as noticed that active haptic interaction gives more satisfaction to the users over passive haptic interaction using Skype.