Whats been happening in the DiVE? Duke Visualization Friday Forum: - PowerPoint PPT Presentation

What’s been happening in the DiVE? Duke Visualization Friday Forum: 4/17/2015 Presented by Regis Kopper + David J. Zielinski 1

DiVE Personnel Joanne Grosshans Regis Kopper Administrative / Booking Director David J. Zielinski Carla Sturdivant Research and Development Finance / Grants Engineer Mark Ogren George A. Truskey Pratt IT Senior Associate Dean for Research 2

DiVE Students Stephanie Fiddy Rebecca Ayana Lai Burkins Trey Elizabeth Emmanuel Bagley Onstwedder Shiferaw Teddy Carolyn Cheng Ward Dugas Ma 3

NSF Major Research Instrumentation Award Proposal Strengths Interdisciplinarity: 14 projects, 3 disciplines, 11 research areas Broader impact : Applicability to different areas, outreach, focus on minorities Two-panel review: BCS ( top 3/19 ) + CNS ( Highly Recommended ) $399,720 Awarded 30% Matching from Duke’s Provost Office $571,029 Total budget for upgrade 4

Overview of “HiDiVE” Upgrade 5

System Install Let’s watch the video! 6

Hardware Upgrade Overview Item 2005 2015 Result Projectors Christie Mirage Christie Mirage Higher Resolution S+2K WU7K-M (x2) Brighter Image Higher Frame Rate GPU Quadro FX 3000 Quadro K6000 Larger Models Quadro FX 5600 Better Lighting/Shaders Shutter Stereographics CE-3 Volfoni Edge RF No line of sight issues Glasses RealD CE-4 (less glasses losing sync) Tracking Intersense IS-900 latest IS-900 Better Tracking Coverage 18 emitters 33 emitters (especially near floor) 7

Projectors 2015 Model Christie Mirage S+2K Christie Mirage Stacked (2) WU7K-M WU7K-M Native 1400x1050 = 1.1 M 1920x1200 1920x1920 = 3.7 M Refresh 120hz (110hz actual) 120hz 120 hz Lumens 3,000 6,300 12,600 8

Overlap/Blending 720 480 Blend Zone 1200 720 1920 Details: ● Projector hardware “Christie Twist” provides alignment and blending. ● NVIDIA Premium Mosaic provides single desktop per wall. 9

Graphics Cards 2015 NVIDIA Quadro FX 3000 NVIDIA Quadro FX 5600 NVIDIA Quadro K6000 Clock 400mhz 600mhz 700mhz Cores n/a 128 2880 Memory 256MB 1.5GB 12GB Tri/Sec 100 million 300 million 2000 million Note: using 2 outputs per K6000 10

Shutter Glasses 2015 RealD CE-4 Volfoni Edge RF Crystal Eyes CE-3 RF IR LIGHT Note: RF has no line of sight issues 11

Tracking System Existing Emitters New Emitters 2005 2015 12

Single Wall System Diagram Machine Room Main DiVE Room Dell Stacked (2) T7400 Christie WU7K-M CAT6 DVI NVIDIA Quadro K6000 Extron DL 201 NVIDIA Quadro Sync 13 To Next Node

Additional VR Hardware (CIEMAS 1411) zSpace 200 VR Workbench/Tabletop Oculus DK1 + DK2 VR HMD Razer Hydra 6DOF Tracking 14

DiVE Software Libraries Library Description License 2005 2015 Avizo sci/med vis Commercial Yes (Amira) Yes Syzygy low level / OpenGL Open Source Yes Yes Virtools game engine Commercial Yes Legacy Unity game engine Commercial No Yes 15

Unity + MiddleVR Plugin Sebastien Kuntz 16

Recent Projects 17

Exploring the Effects of Image Persistence in Low Frame Rate Virtual Environments Regis Kopper David J. Zielinski Hrishikesh M. Rao Marc A. Sommer Duke immersive Virtual Environment Duke immersive Virtual Environment Dept. of Biomedical Engineering Dept. of Biomedical Engineering Duke University Duke University Duke University Duke University 18

Low Persistence - Explanation Issues: HP multiple “phantom lines” artifacts for moving objects, LP has strobing effect. 19

Low Persistence - Selection Task HFR significantly faster (p < .0001) than HP and LP No significant difference between HP and LP. 20

Low Persistence - Navigation Task Path Width vs Display Method HFR significantly faster than LP (p < .05) and HP (p < .005) 21

Low Persistence - Future Work ● Test targets in motion. ● Collaborate with professor Greg Appelbaum. 22

Perceptual Load in the Kitchen: Measuring Distraction in Immersive Virtual Reality Bettina Olk Regis Kopper David J. Zielinski Jacobs University Duke University Duke University K Perceptual Load Theory: X M Z N O O N H X O O Distraction is stronger when W O perceptual load is low, rather Low High then high. Perceptual Perceptual Load Load 23

Perceptual Load - Experiment Incongruent Congruent Low Load (target=soda) High Load (target=yogurt) ● User searched for target within the circle (soda or yogurt) ● Flanker (congruent or incongruent) had to be ignored. ● Perceptual load was manipulated by the color of the distractors. 24

Perceptual Load - Results ● Contrary to theory, distractor effect not modulated by load. ● Use VR to test traditional research, in complex ecologically valid scenarios. 25

Duke Health and Wellness Center Vis Jun Yin John H Nicholson Student Professor Master of Engineering Master of Engineering Managment Management A builder and developer of some of the most highly visible projects domestically and internationally. Pre-Visualization of Examination Room 26

Structual Engineering Joe Nadeau Professor Civil Engineering CEE429: Final Projects Next Week 27

Wayfinding by Auditory Cues in Virtual Environments Regis Kopper Ayana Burkins Duke University CS Junior Duke University Wayfinding: cognitive component of navigation Hypothesis: spatial audio will aid in spatial memory tasks 28

Auditory Cues - Environment ● Two mazes ● With and without spatial sound ● Visible landmarks 29

Auditory Cues - Pilot Evaluation Travel to target With Sound Without Sound 30

Dig@IT: Virtual Reality in Archaeology Emmanuel Cheng Ma Nicola Lercari Maurizio Forte Regis Kopper Shiferaw Duke University UC Merced Duke University Duke University Duke University ● Archaeology preservation ● Scanning of multiple excavation phases ● Multi platform (Currently implemented on Oculus Rift) ● Dynamic model loading Çatalhoük archaeological site in Turkey 31

Dig@IT Features Digital Archaeology Timeline Tool In-Context Menu Tablet Tool 32

Dig@IT Let’s watch the movie ! 33

Comparison of Interactive Environments for the Archaeological Exploration of 3D Landscape Data Rebecca Bennett Regis Kopper David J. Zielinski University of Winchester Duke University Duke University Remote sensing for archaeological landscapes 34

Landscape Archeology - Vis Layers Different observations based on visualization layer Aerial photograph Elevation map More history can be seen via Lidar. Local relief model Hillshade 35

Landscape Archeology - User Study 3D web application DiVE 2D GIS application ● 6 Expert participants ● 6 Inexperienced participants ● Ability to change visualization layer ● Think-aloud protocol 36

Landscape Archeology - Results 37

How to Get Involved ● Taking courses ● Work study ● Volunteering ○ tour guides ○ projects ○ website development ● Collaborations ○ Projects that can benefit from immersive vis ○ Opportunities for grant funding 38

ISIS 320 - Introduction to Interaction Design and Virtual Environments ● Spring 2016 Course ● Concepts of Human-Computer Interaction and Virtual Reality Technology ● User-centered design applied to 3D user interfaces ● Project-oriented course May be taken as CS independent study! 39

DiVE Public Open Houses Tuesdays 4-5pm Experience DiVE apps First come first served regis.kopper@duke.edu djzielin@duke.edu Questions? 40