Stanford 50: State of the Art and Future Directions of Computational Mathematics and Numerical Computing


  • March 31, 2007
  • 11:25 am - 11:50 am

    Duality methods for nonlinear image processing

    Tony Chan (UC Los Angeles)

    I review how a primal-dual algorithm for minimizing the total variation norm that Gene and I developed with Pep Mulet a decade ago is related to recent work by A. Chambolle on dual projection algorithms and by W. Yin and D. Goldfarb on second-order cone algorithms for the same problem.

    Stanford University Home Page