From Simon Julier: "The skew simplex form is actually a bad set of points to use (the paper for 2003 ACC replaces them with something much more sensible) but the scaling paper (which deals with manipulating the higher order properties of the distribution) uses the skew one for context."

@INPROCEEDINGS{inproceedings:julier02,
  Address = {Anchorage AK, USA},
  Author = {S. J. Julier{, and J. K. Uhlmann}},
  Booktitle = {Proceedings of the IEEE American Control Conference},
  Month = {8--10 May},
  Pages = {887--892},
  Publisher = {IEEE},
  Title = {{Reduced Sigma Point Filters for the Propagation of Means and Covariances through Nonlinear Transformations}},
  Year = {2002}
}
@INPROCEEDINGS{inproceedings:julier02a,
  Address = {Anchorage AK, USA},
  Author = {S. J. Julier{, and J. K. Uhlmann}},
  Booktitle = {Proceedings of the IEEE American Control Conference},
  Month = {8--10 May},
  Pages = {4555--4559},
  Publisher = {IEEE},
  Title = {{The Scaled Unscented Transformation}},
  Year = {2002}
}

Return to Greg Welch's Kalman filter page.
Return to Greg Welch's home page.

Last modified: Wednesday, March 5, 2003