You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Problem Description:
The initiate method of MultiMeasurementInitiator currently assumes that the associations provided by the initiator's data_associator will be SingleHypothesis types. However, this will error when the user configures the initiator to use a hypothesiser that returns MultipleHypothesis types as it does not define measurement or prediction members.
Steps to Reproduce:
Define a tracker with a MultiMeasurementInitiator that is using a PDAHypothesiser within it's data_associator.
Desired Solution:
Ideally, the MultiMeasurementInitiator should be able to use any valid hypothesiser/data_associator combination, not just ones that can produce SingleHypothesis types.
It may not make sense for the MultiMeasurementInitiator itself to work in a Multi-Hypothesis context (especially with the releasability criteria being number of detections). At the very least, this should be mentioned in the documentation.
In the case of 2, it might make sense to implement an analogous initiator that can specifically release tracks based on some sort of analysis of the MultipleHypothesis type (like how the MixtureTracker's __next__ method is implemented).
The text was updated successfully, but these errors were encountered:
Problem Description:
The
initiate
method ofMultiMeasurementInitiator
currently assumes that the associations provided by the initiator's data_associator will beSingleHypothesis
types. However, this will error when the user configures the initiator to use a hypothesiser that returnsMultipleHypothesis
types as it does not definemeasurement
orprediction
members.Steps to Reproduce:
Define a tracker with a
MultiMeasurementInitiator
that is using aPDAHypothesiser
within it'sdata_associator
.Desired Solution:
MultiMeasurementInitiator
should be able to use any valid hypothesiser/data_associator combination, not just ones that can produceSingleHypothesis
types.MultiMeasurementInitiator
itself to work in a Multi-Hypothesis context (especially with the releasability criteria being number of detections). At the very least, this should be mentioned in the documentation.MultipleHypothesis
type (like how theMixtureTracker
's__next__
method is implemented).The text was updated successfully, but these errors were encountered: