Multitarget Data Association with Higher-Order Motion Models

Abstract

We present an iterative approximate solution to the multidimensional assignment problem under general cost functions. The method maintains a feasible solution at every step, and is guaranteed to converge. It is similar to the iterated conditional modes (ICM) algorithm, but applied at each step to a block of variables representing correspondences between two adjacent frames, with the optimal conditional mode being calculated exactly as the solution to a two-frame linear assignment problem. Experiments with ground-truthed trajectory data show that the method outperforms both network-flow data association and greedy recursive filtering using a constant velocity motion model.

Cite

Text

Collins. "Multitarget Data Association with Higher-Order Motion Models." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2012. doi:10.1109/CVPR.2012.6247870

Markdown

[Collins. "Multitarget Data Association with Higher-Order Motion Models." IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2012.](https://mlanthology.org/cvpr/2012/collins2012cvpr-multitarget/) doi:10.1109/CVPR.2012.6247870

BibTeX

@inproceedings{collins2012cvpr-multitarget,
  title     = {{Multitarget Data Association with Higher-Order Motion Models}},
  author    = {Collins, Robert T.},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year      = {2012},
  pages     = {1744-1751},
  doi       = {10.1109/CVPR.2012.6247870},
  url       = {https://mlanthology.org/cvpr/2012/collins2012cvpr-multitarget/}
}