Geometric Modeling under Uncertainty

21

  • We have studied various aspects of geometric modeling. Recently, we use machine learning techniques to handle uncertainty in geometric modeling. We typically use probabilistic graphical models to represent geometric variables tied with observed values.
  • Recent research projects are :
    • 3D Reconstruction from Freehand Ultrasound Images
    • Combining volumetric CT and optical scan data

 

Stochastic Optimization

3

  • We are also interested in stochastic optimization problems such as stochastic approximation and simulation optimization. techniques.
  • Recent research projects are :
    • Global root finding with Gaussian Process Regression
    • Simulation optimization for military RAM problem