Horizon CDT Research Highlights

Research Highlights

Brain Controlled Adaptive Virtual Reality Horror Experiences

  Callum Berger (2021 cohort)

The aim of this research is to understand the application of brain activity data as dynamic feedback into virtual reality horror experiences to personalise the experience based on user’s fears, where dynamic refers to the data being used in real-time to affect the mechanics of the experience. This includes creating a unique experience for everyone by learning the individual’s fears and exposing them to these fears within a controlled and immersive experience. Breaking this down into sectors compiles a series of research goals of which to be achieved throughout this PhD. These sectors include the understanding of brain data and its relation to fear, the process of adapting the experiences with relation to the direction of narrative applied, as well as the implementation of the experiences within virtual reality and the approach to immersion and design from a performance perspective.

Brain activity data will be retrieved using BCI (Brain-Computer Interface) devices, where devices will be tested for effectiveness such as temporal and spatial resolution. Additionally, fear is stated to be begin within the Amygdala in the temporal lobe. This means any devices used to retrieve fear data must be capable of sensing and outputting this activity. Current potential devices include the emotive insight and the NextMind which gather electroencephalogram (EEG) data, however other types such as functional near-infrared spectroscopy (fNIRS) will also be used to test for capability when sensing fear. To ensure that the data being retrieved is correlated to fear, research will need to be undertaken to understand the data retrieved from these devices and whether fear can be detected. This will include the use of the pilot project to gather some initial brain activity data for analysis as well as research on similar projects that have claimed to have retrieved fear from brain-computer interface (BCI) devices. Potential routes for analysis include using machine learning or fuzzy logic to perform the data analysis before being applied as dynamic feedback.

Adaptation and narrative provide a much broader approach with many possibilities providing different end goals. Research will be conducted into the ways in which adaptation can be applied, such as diverging narratives in response to a user’s reaction to a fear. An example of this would be a narrative focused on a fear of spiders due to a detected fearful reaction from spiders. An issue that could occur from this type of routed narrative is the creation of an enclosed narrative to a specific fear, therefore routes would need to be inclusive of multiple fears and ensure that a user does not get exposed too much to one single fear but instead a range that can provide a more personalised unique experience. Designing the experiences relies heavily on the atmosphere and the setting decided. Immersion is important to maintaining realism when within a virtual environment and therefore will be key to keeping users engaged within the experiences.