Bayesian Filtering Library 0.6.0 released

The Bayesian Filtering Library development team is pleased to announce the 0.6.0 release of BFL.
You can download this release from here and read the installation instructions online (also reachable through the orocos website).

This release includes support for lti, boost and newmat as matrix library and lti and boost as random number generator.
A new feature is the backward filter and smoother algorithm and the CPPUnit tests.
Furthermore for the first time, a step-by-step installation guide is available for Visual Studio on Windows.

In detail this release addresses the following reported issues:

  ID            Summary 
  303    The future of BFL (aka: BFL needs new maintainer) 
  319    add backward filter and tests to build system 
  320    Default implementation for virtual functions 
  321    const function arguments in mcpdf class 
  329    Add function to get one sample + change int into unsigned... 
  330    Sample::ValueSet() does not adjust dimension 
  331    BFL should use return codes or c++ exceptions 
  333    Sample stores dimension 
  334    No need to re-implement virtual functions 
  335    Cleanup of some pdf code 
  343    PostGet() should return a more specific Pdf if possible 
  349    Add SVN revision number to doxygen generated docu 
  350    make analytic system and measurement model consistent 
  351    Extension for IteratedExtendedKalmanFilter 
  389    Examples refuse to compile 
  392    Change build system to cmake 
  393    Not possible to build static libraries 
  395    Automate building of Ubuntu/Debian packages 
  400    Cholesky decomposition 
  403    Building BFL in Windows 
  411    Boost needs pinv implementation 
  416    License issues for BFL template code

Details are available through: this link

The Bayesian Filtering Library (BFL) provides an application independent framework for inference in Dynamic Bayesian Networks, i.e., recursive information processing and estimation algorithms based on Bayes' rule, such as (Extended) Kalman Filters, Particle Filters (or Sequential Monte Carlo methods), etc. These algorithms can, for example, be run on top of the Realtime Services, or be used for estimation in Kinematics & Dynamics applications.