A Consistent Metric for Performance Evaluation of Multi-Object Filters

By Dominic Schuhmacher, B. Vo, Ba-Ngu Vo. IEEE Transactions on Signal Processing, 2008

2,077 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Civil and Structural Engineering, Computer Networks and Communications

View Paper: https://www.semanticscholar.org/paper/22b26297e0cc5df3efdba54a45714e4e27b59e17

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Civil and Structural Engineering

In Computer Networks and Communications

Citations in the corpus

Clustering for point pattern data (2017, background): “However, it does not have a physically consistent interpretation when the point patterns have different cardinalities, see [31] for further details.”
Marginal multi-bernoulli filters: RFS derivation of MHT, JIPDA, and association-based member (2012, methodology): “5; the figure shows the mean optimal sub-pattern assignment (OSPA) 17 [40] with p = 1 and c = 20 (calculated over both position and velocity) as a function of time, averaged over 200 Monte Carlo trials.”
A phd Filter for Tracking Multiple Extended Targets Using Random Matrices (2012, background): “The results are presented in terms of the multi-target measure optimal subpattern assignment metric (OSPA) [29], cardinality and length of the estimated major and minor axes of the extension matrices.”
MOT20: A benchmark for multi object tracking in crowded scenes (2020, background): “In the past, a large number of metrics for quantitative evaluation of multiple target tracking have been proposed [2], [6], [12]–[14], [16].”
High-Accuracy Localization for Assisted Living: 5G systems will turn multipath channels from foe to friend (2016, methodology): “It has been accomplished in [18] and [33] using a constrained optimal subpattern assignment approach [35].”
Multi-leaf tracking from fluorescence plant videos (2014, background): “There are prior work in evaluating multi-target tracking performance [22] [23] .”
A Random-Finite-Set Approach to Bayesian SLAM (2011, background): “9 shows that the computation time is comparable with that of the MHT-FastSLAM algorithm, both of which are less expensive than JCBB–EKF–SLAM as its hypothesis tree grows in the presence of high clutter.”
Labeled Random Finite Sets and Multi-Object Conjugate Priors (2013, positive): “4 compares the OSPA miss distance ( m) [33] for the two filters.”
Consensus CPHD Filter for Distributed Multitarget Tracking (2013, methodology): “Multitarget tracking performance is evaluated in terms of the OSPA (Optimal SubPattern Analysis) metric [41].”
Joint Detection and Estimation of Multiple Objects From Image Observations (2010, methodology): “We use the optimal subpattern assignment (OSPA) distance between the estimated and true multi-object state as the estimation error, since it jointly captures differences in cardinality and individual elements between two finite sets in a mathematically consistent yet intuitively meaningful way [ 19 ”
An Efficient Implementation of the Generalized Labeled Multi-Bernoulli Filter (2016, background): “1) Case 1: This baseline comparison uses the values of H where each implementation starts to exhibit reasonable tracking performance, and have approximately the same average optimal sub-pattern assignment (OSPA) error [55].”
On performance evaluation of multi-object filters (2008, background): “This article presents a summary of results, full details can be found in the paper [9].”
MOT16: A Benchmark for Multi-Object Tracking (2016, background): “In the past, a large number of metrics for quantitative evaluation of multiple target tracking have been proposed [8], [30], [38], [40], [41], [46].”
Multisensor data fusion: A review of the state-of-the-art (2013, methodology): “Aside from the conventional approaches for performance measurement, there is some notable work on development of MOPs for multi-target fusion systems within the Finite Set theory framework [188,189].”
A Metric for Performance Evaluation of Multi-Target Tracking Algorithms (2011, methodology): “[8] demonstrated the OSPA metric in the context of performan ce evaluation of multi-target filtering algorithms.”
MOTChallenge 2015: Towards a Benchmark for Multi-Target Tracking (2015, background): “Even though many measures have been proposed in the past [12], [33], [44], [46], [47], [52], comparing a new method against prior art is typically not straightforward.”
Labeled Random Finite Sets and the Bayes Multi-Target Tracking Filter (2013, background): “Figures 8 and 9 show the OSPA distance [51] and its localization and cardinality components forc = 100m andp = 1.”