Qi Mo

 

Home

Research

Publications

Courses

Links

COMP 768 Project Proposal

Motivation

Simulating hair is very important for realistic simulation and rendering of virtual human, however, it remains a challenge especially for interactive applications. Although rencently animation industry have been able to produce highly plausible hair with accurate motion and appearance, these results were mostly obtained offline. The biggest challenge that has been perventing hair simulation from achieving real-time is the large number of individual hair strands that need to be simulated and rendered, the average number of a full head of hair consists of no less than 100,000 strands. 

Various techniques have been proposed to accelerate hair simulation, either by using simplified geometry, multiresolution geometrical representaiton and simulation LODs, or by using graphics hardware. However, over-simplification might lead to hair appearing too chunky or unnatural, especially for curly hair, and there might be visible artifacts between transtion of different level of details. These limitations mean that we are still facing the challenge of real-time hair simulation, and this motivates this project.

Literature and Background

There have been three different frameworks of multiresolution representations of hair. One is the wisp model. The authors started from addressing the static modeling problem and went further to handle the dynamics by building a separate simulation LOD callsed adaptive wisp trees. Another proposed multiresolution approach chose to use three levels, strip, cluster, and strand, each based on subdivided surfaces and curves. It incorporates simulation LODs and is able to do continuous transitions dynamically on the fly. A more recent model used super-helices to model hair strands and used the derived dynamic equaitons to handle the dynamics. This is a more physically accurate model but is also more computationally demanding, therefore the authors were currently limited to only using ten guided strands to model a full head of hair. 

Poposed Tasks and Goals

A short term goal of this project is to migrate Kelly Ward's hair LOD model to GPU. This part overlaps with one of my conceurrent projects that implements SSVs on GPU, because SSV is the underlying BV used for Kelly's hair LODs. A long term goal is to incorporate and improve in the aspect of simulation LODs. In particular, current transition criteria do not take collision into account, which might be a major shortcoming. A more remote goal would be to try to apply LOD technique on the more physically-based model of Kerchhoff equations and super helices.

Timeline and Checkpoints

By April 7th, I aim to have implemented a basic geometric LODs using graphics hardware, including collision detection and simple mass spring system dynamics. By May 5th, I aim to have implemented more improved dynamics system with simulation LODs.

Project Proposal Slides

           Reference

        

          1. T.-Y. Kim and U. Neumann, “Interactive multiresolution hair modeling and editing,” ACM Transactions on Graphics, vol.                   21, no. 3, pp. 620– 629, July 2002, proceedings of ACM SIGGRAPH 2002.

            2. K. Ward and M. C. Lin, “Adaptive grouping and subdivision for simulating hair dynamics,” in Pacific Graphics Conference                   on Computer Graphics and Applications, October 2003, pp. 234–243.

            3. F. Bertails, T.-Y. Kim, M.-P. Cani, and U. Neumann, “Adaptive wisp tree -a multiresolution control structure for simulating                 dynamic clustering in hair motion,” in ACM SIGGRAPH Symposium on Computer Animation, July 2003, pp. 207–213.

            4. F. Bertails, B. Audoly, M.-P. Cani, B. Querleux, F. Leroy, and J.-L. Lev.eque, "Super-helices for predicting the                           dynamics of natural hair," in ACM Transactions on Graphics (Proceedings of the SIGGRAPH conference), August 2006.

            5. K. Ward, M. C. Lin, J. Lee, S. Fisher, and D. Macri, “Modeling hair using level-of-detail representations,” in International                   Conference on Computer Animation and Social Agents, May 2003, pp. 41–47.

            6. K. Ward, F. Bertails, T.-Y. Kim, S. R. Marschner, M.-P. Cani, M. C. Lin. A Survey on Hair Modeling: Styling, Simulation                 and Rendering. IEEE Transactions on Visualization and Computer Graphics, vol. 13, no. 2, pp. 213-234, Mar/Apr. 2007.

            7. E. Plante, M.-P. Cani, P. Poulin. "A Layered Wisp Model for Simulating Interactions inside Long Hair." in Proceedings of                   Eurographics Worshop on Animation and Simulation. 2001.

            8. K. Ward, N. Galoppo, M. Lin. "Interactive Virtual Hair Salon." Presence. 2007. Vol. 16. No. 3.

            9. K. Ward. "Simplified Representatiins for Modeling Hair." UNC-CS Tech report. 2002.