A discrete chain graph model for 3d+t cell tracking with high misdetection robustness

Authors:
Bernhard X. Kausler;Martin Schiegg;Bjoern Andres;Martin Lindner;Ullrich Koethe;Heike Leitte;Jochen Wittbrodt;Lars Hufnagel;Fred A. Hamprecht
Affiliations:
HCI/IWR, Heidelberg University, Germany;HCI/IWR, Heidelberg University, Germany;HCI/IWR, Heidelberg University, Germany,SEAS, Harvard University, United States;HCI/IWR, Heidelberg University, Germany;HCI/IWR, Heidelberg University, Germany;HCI/IWR, Heidelberg University, Germany;COS, Heidelberg University, Germany;European Molecular Biology Laboratory (EMBL), Heidelberg, Germany;HCI/IWR, Heidelberg University, Germany
Venue:
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part III
Year:
2012

Citing 8
Cited 0

Random Forests

Machine Learning
Object tracking: A survey

ACM Computing Surveys (CSUR)
An empirical comparison of inference algorithms for graphical models with higher order factors using openGM

Proceedings of the 32nd DAGM conference on Pattern recognition
Cell image analysis: Algorithms, system and applications

WACV '11 Proceedings of the 2011 IEEE Workshop on Applications of Computer Vision (WACV)
Markov Random Fields for Vision and Image Processing

Markov Random Fields for Vision and Image Processing
Multiobject tracking as maximum weight independent set

CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
A tutorial on particle filters for online nonlinear/non-GaussianBayesian tracking

IEEE Transactions on Signal Processing
Factor graphs and the sum-product algorithm

IEEE Transactions on Information Theory

Quantified Score

Hi-index	0.00

Visualization

Abstract

Tracking by assignment is well suited for tracking a varying number of divisible cells, but suffers from false positive detections. We reformulate tracking by assignment as a chain graph---a mixed directed-undirected probabilistic graphical model---and obtain a tracking simultaneously over all time steps from the maximum a-posteriori configuration. The model is evaluated on two challenging four-dimensional data sets from developmental biology. Compared to previous work, we obtain improved tracks due to an increased robustness against false positive detections and the incorporation of temporal domain knowledge.