Multi-interface gesture based organic modelling Bradley Wesson
Overview Background Rationale Development Issues Testing Conclusion and Future Work
Background – The problem “Designers know more than they can say” (Schön, 1992) “[creative interfaces] should capture users’ intended ambiguity, vagueness, and imprecision and convey these qualities visually and through interactive behavior” (Gross, 1996) “computer models […] are typically considerably more difficult to create [than with pen and paper].” (Zeleznik, et al., 2006)
Background – Current solutions ILoveSketch (Bae, Balakrishnan, & Singh, 2008) FiberMesh (Nealen, Igarashi, Sorkine, and Alexa, 2007)
Project Combine Kinect with Mobile for more natural interactions Natural gestures Virtual “tool” Buttons and orientation sensing
Project – Interactions Kinect Right hand for interactions Clench fist to grab geometry Both hands to reorient model Left hand to change size Kinect+Mobile Virtual tool Physical button to grab geometry Two fingers on screen to reorient Left hand to change size
Project - Demo
Project – Mobile Device orientation, buttons, touch screen TCP connection over Ad Hoc network Send 20 packets per second Prevent flooding network
Project – The model Stored as voxels Valve 0.5-depth threshold Voxels “smudged” Marching Cubes: A high resolution 3D surface construction algorithm (Lorensen & Cline, 1987) Improved Alpha-tested Magnification for Vector Textures and Special Effects (Green, 2007)
Project – Rendering Depth perception Lighting, transparency
Issues Trouble determining depth of interaction relative to geometry Inaccuracies of mobile sensors make linear acceleration unreliable Poor development platform choice Translation Acceleration Velocity
Timeline Testing Thesis Expo Rslts Seminar Mid-project Ethics Development Lit Review Proj Proposal MayMarchJanuaryNovemberSeptemberJulyMayMachJuly
Timeline - Revised Testing Thesis Expo Rslts Seminar Mid-project Ethics Development Lit Review Proj Proposal MayMarchJanuaryNovemberSeptemberJulyMayMachJuly C++ Prototype C# Rebuild XNA Port Always finding new papers!
Testing / Measurements Modelling a teapot Measurements Observation Survey NASA TLX
Evaluation Quality of product Speed of execution Measured workload Perceived ease of use Perceived speed Enjoyment I’m not using that!
Conclusion Development process Accuracy Built around creative expression Multi-interface testing Future work Implement Kalman filter for linear acceleration Painting and texture Animation
Questions ?