Wrong data association used

[cdondrup]

There are three different forms of data association:

• NN
• JPDA
• NNJPDA

A quick look at https://github.com/LCAS/bayestracking/blob/master/include/bayes_tracking/multitracker.h#L220 and https://github.com/LCAS/bayestracking/blob/master/include/bayes_tracking/multitracker.h#L263 suggests that actually JPDA is used if the user specifies that NNJPDA should be used. Since the JPDA case is not covered the tracker will out put ###### Unknown association algorithm: 1 ##### if JPDA is selected.

If I am right a simple change of https://github.com/LCAS/bayestracking/blob/master/include/bayes_tracking/multitracker.h#L220 to if (alg == JPDA) should do the trick.

[cdondrup]

Actually it doesn't seem to be that easy. @nbellotto could you please have a look why and where JPDA is missing?

Must be somewhere in this block:

if (alg == NN) {  /// NN data association
        amat.computeNN(CORRELATION_LOG);
        // record unmatched observations for possible candidates creation
        m_unmatched = amat.URow;
        // data assignment
        for (size_t n = 0; n < amat.NN.size(); n++) {
            m_assignments.insert(std::make_pair(amat.NN[n].row, amat.NN[n].col));
        }
      }
      else if (alg == NNJPDA) { /// NNJPDA data association (one sensor)
        // compute associations
        jpda->getAssociations();
        jpda->getProbabilities();
        vector< jpda::Association > association(znum.size());
        jpda->getMultiNNJPDA(association);
//         jpda->getMonoNNJPDA(association);
        // data assignment
        jpda::Association::iterator ai, aiEnd = association[0].end();
        for (ai = association[0].begin(); ai != aiEnd; ai++) {
          if (ai->t) {   // not a clutter
            m_assignments.insert(std::make_pair(ai->z, ai->t - 1));
          }
          else {   // add clutter to unmatched list
            m_unmatched.push_back(ai->z);
          }
        }
                                                                  delete jpda;
      }
      else {
        cerr << "###### Unknown association algorithm: " << alg << " #####\n";
        return false;
      }

[marc-hanheide]

seems JPDA is not really meant to be used: https://github.com/LCAS/bayestracking/blob/master/include/bayes_tracking/multitracker.h#L42

[cdondrup]

This is why I "fixed" it so you can't choose it in the config file any more when using the people tracker. Not sure why this is not supposed to be used, would have to ask @nbellotto