VR (Virtual Reality) technology has been widely used in various application areas ranging from entertainment such as game, movie to education, medical, manufacturing industry. However, it is found that VR users can experience severe fatigue and symptoms of motion sickness, which is referred to as cyber sickness. It is a serious obstacle of such technology and growth of the VR market.
Various companies and research institutes are currently developing the solutions to predict and reduce VR-sickness but there is no practical, systematic solution yet.
A Korean R&D institute focusing on ICT Bio-healthcare has developed highly reliable VR sickness monitoring and analysis technology combining with the elements of VR contents based on the data extracted from large-scale clinical data of more than 500 people. The technology is approved by international standard, IEEE-SA WG 3079 in 2020.
The functions performed by the technology as below.
1) Select a section to predict the level of sickness in VR video
2) Predict the level of sickness quantitatively within the selected section
3) Visualise the graph of sickness level with a real value (0~5)
4) Save and load the data of sickness prediction
The technology has been developed as a Unity Plug-in, meaning that VR developers can use the software for VR contents creating immediately. A Korean game company has been already launched new VR game using the technology.
Any companies, research institutes or universities creating VR contents or providing VR services in entertainment, sports, education, medical, healthcare or defense industry can improve their VR contents to be more acceptable to users. Also, the Korean R&D institute is interested in applying joint R&D programme (e.g. EUREKA, EUROSTARS) to upgrade the software speed and system optimization, etc.
Advantages & innovations
It is a highly reliable VR sickness analysis technology realized through machine learning processing of large-scale empirical data acquired by extensive clinical tests on users, and its features are as follows:
- Exploits the features extracted from large-scale clinical-test data on more than 500 people
- In-depth analysis of various categorized data via machine learning: the parametrized characteristics of the given VR contents (camera/object motions, complexity, depth, etc.) bio-signal (EEG(Electroencephalogram), pulse wave and GSR(Galvanic Skin Reflex) and human motions (eye-tracking, head-tracking, etc.)
- Above 89% of reliability factor in terms of VR-sickness prediction accuracy
- Quantified prediction of VR sickness via machine learning model of human perception process
- Interpretation of users’ susceptibility response in terms of bio-metrics, called biomarkers
- Adoption and approval of the VR sickness analysis and quantification methodology by the IEEE international standardization organization in 2020
Stage of development
Already on the market
Partner sought
- Type of partner sought: SMEs or research institute
- Specific area of activity of the partner: VR contents services in entertainment, sports, education, medical, healthcare, defense industry
- Task to be performed: Employing the technology to improve their VR contents to be more acceptable to users. Applying to joint R&D programme(e.g. EUREKA, EUROSTARS) to upgrade the software speed and system optimization, etc.