Introducing
Your new presentation assistant.
Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.
Trending searches
VR Metrics - Concept
40
VR Metrics
Concept Proposal
Natnael Ghilamicael
Georgios Nurcan-Georgiou
Mario Sagawa
Naomi Sagawa
VR METRICS
How can we track and measure the amount of movement in VR?
BRAINSTORMING
- initial research
- discussion on possible focal points
- phrasing of possible research questions
- quantitative collection of ideas and material
- state of the art
Brainstorming
FOCAL POINTS
- evaluate & compare existing methods
- focus on health aspect (posture, strain etc.)
- combine different VR movement measurements
- find a new way to track movement and exertion
- implement exergame applying movement metrics
MOVEMENT VS EXERTION
MOVEMENT
EXERTION
- native VR usually capable of some sort of movement measurement
- spatial detection of one or several points - translation into digital environment
- movement of different body parts or muscle groups
- 3-dimensional measurement
- generally challenging to measure[2]
- heart rate monitoring
- subjective rating
- Borg scale/RPE scale[1]
- lactate level in breathing gas
- estimate burned calories
- combination of several measurements + movement
research questions
[1] https://www.cdc.gov/physicalactivity
[2] https://www.mayoclinic.org/
STATE OF THE ART
[1]
[2]
[3]
[4]
- motion capturing suits
- motion capturing markers
- camera-based movement tracking
- 2D pose estimation
- 3D pose estimation with depth cameras
- exertion tracker
- additional wearable devices
[5]
[7]
[9]
[6]
[8]
[1] https://www.deepmotion.com/
[2] https://teslasuit.io/
[3] https://optitrack.com/
[4] https://yur.fit/
[5] https://vrhealth.institute/vr_exercise_tracker/
[6] Cao et al., OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields, arXiv:1812.08008
[7] https://www.natml.ai/
[8] https://www.rokoko.com/products/smartsuit-pro
[9] https://developer.microsoft.com/en-us/windows/kinect/
CONCRETIZATION
Concretization
- focus: find non-intrusive and affordable method to track movement and estimate exertion
- concentrate on 2 ideas from initial brainstorming
- weigh up pros and cons of ideas
- take into consideration usefulness and feasability
BIO-SENSING VEST
bio-sensing tracking vest
+ innovative
+ possibly very accurate
+ light weight
+ encapsulation
- research value
- wearable device
- worn throughout
whole session
- encapsulation
- wearable device in form of light weight vest worn on top of regular clothing
- equipped with stretch sensors to track breathing patterns and heart rate
- additional belt to track core stability and muscle activation
- independent from actual VR application and system
- data evaluation via arduino
- comparison of results to data collected using 'proven methods'
CAMERA-BASED TRACKING
- usage of one or several cameras to track person and estimate posture
- group larger muscle group
- visually track muscle movement
- calculate exact movement
- estimate exertion
Can a simple camera-based tracking system give accurate and sufficient information about movement and exertion in VR?
enhanced camera-based tracking system
SZENARIO
szenario
- initial calibration phase
- visual tracking and mapping of the body onto a simplified model
- division of the large muscle groups
- additional wearable device to measure heart rate and breathing patterns
- actual tracking
- reduced onto camera-based system only
- working with data collected during calibration phase
- evaluation
- evaluating data (possibly in real-time)
- estimating exertion via different parameters
CONCEPTUALIZATION
Can a simple camera-based tracking system give accurate and sufficient information about movement and exertion in VR?
Conceptualization
CONCEPT
camera-based tracking system
evaluation &
application
calculation
calibration
Concept
estimation
tracking
Technologies to Explore
- VR Health Exercise Tracker: gather baseline data[1]
- YUR Fitness and Health Tracker for VR: alternative[2]
- OpenPose: pose estimation[3]
- MoveNet3D: pose estimation[4]
- depth-camera vs multiple 2d cameras
- wearable devices for measuring heart rate and breathing patterns
- fitness wristband
- breast belt
Can a simple camera-based tracking system give accurate and sufficient information about movement and exertion in VR?
[1] https://vrhealth.institute/vr_exercise_tracker/
[2] https://yur.fit/
[3] Cao et al., OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity
Fields, arXiv:1812.08008
[4] https://www.natml.ai/
TIMELINE
evaluate
experiment
- evaluate framework
- possibly build small VR exergame for testing
- conclude success
- experiment with different setups
- evaluate and compare different technologies
- cameras
- tracking devices
- tracking software
Timeline
develop framework
gather data
- build a framework based on the findings
- focus on system being easy-to-use
- plug-and-play mechanic
- conduct a study using the known methods of VR tracking
- collect base-line data
- VR health exercise tracker
RISKS
& FALLBACK STRATEGIES
Risks
no real-time processing possible
- separate data processing from VR application
- allow time delayed processing
no accurate results
- usage of additional sensors (at major joints)
- keep simple wearables (fitness tracker)