An Affective Interaction System using Virtual Reality and Brain-Computer Interface.

Area of Science:

• Affective Computing
• Human-Computer Interaction
• Neuroscience

Background:

• Affective Computing aims to enable computers to recognize human emotions.
• Applications span healthcare, gaming, and intuitive interface design.
• Existing systems often rely on complex or expensive equipment.

Purpose of the Study:

• To propose and evaluate a novel affective interaction system integrating dry EEG-based Brain-Computer Interface (BCI) and Virtual Reality (VR).
• To assess the system's capability in recognizing positive and negative emotions using low-cost consumer devices.
• To demonstrate the feasibility of real-time affective interaction through BCI-VR.

Main Methods:

• Developed a BCI-VR system using a consumer-grade EEG headband with dry electrodes and a VR headset with an Android smartphone.
• Acquired EEG signals wirelessly and processed them using in-house software to elicit emotional responses via VR content.
• Extracted EEG bandpower features to train Linear Discriminant and Support Vector Machine classifiers.

Main Results:

• The proposed BCI-VR system achieved approximately 66% accuracy in classifying positive versus negative emotions on a custom dataset.
• Performance on the custom dataset was comparable to results obtained on the public SEED-IV dataset, which used wet EEG electrodes.
• Demonstrated that low-cost, dry EEG-based BCI-VR systems can effectively detect basic emotional states.

Conclusions:

• The developed low-cost BCI-VR system shows significant promise for real-time affective interaction applications.
• Dry EEG electrodes offer a viable alternative to wet electrodes for emotion recognition in BCI-VR systems.
• This research paves the way for more accessible and widespread use of affective computing technologies.