Graph Solvers

Graph Solvers#

Supported Solvers#

NavAbility provides both parametric and nonparametric (i.e. non-Gaussian / multi-modal) factor graph solving solutions. Please reach out via email info@navability.io or Slack if you would like to learn more about these features.

Supported Probabily Models#

Gaussian Case#

Parametric solving supports measurement models using the probability beliefs:

  • Normal

  • MvNormal

And also currently in beta:

  • Max-Mixture of Normal or MvNormal

Non-Gaussian Case#

Non parametric solving can support a much wider range of measurement probability models. Currently the SDK supports

  • Mixture

  • Uniform

  • Rayleigh

  • as well as multihypothesis features per factor – i.e. Categorical.

Many more probability types are natively supported by the solver and yet directly exposed through the SDK, including

  • ManifoldKernelDensities,

  • Heatmaps or Intensity maps,

  • Levelsets,

  • Most common parametric probability models.

Reach out to NavAbility if you seek earlier SDK support for these or other more advanced probability features.

Probability Model Index#

Contents