Title

(Mp) 2T: Multiple People Multiple Parts Tracker

Keywords

body part tracking; humans; k-shortest paths; multi-target; network flow optimization; pedestrians; tracking

Abstract

We present a method for multi-target tracking that exploits the persistence in detection of object parts. While the implicit representation and detection of body parts have recently been leveraged for improved human detection, ours is the first method that attempts to temporally constrain the location of human body parts with the express purpose of improving pedestrian tracking. We pose the problem of simultaneous tracking of multiple targets and their parts in a network flow optimization framework and show that parts of this network need to be optimized separately and iteratively, due to inter-dependencies of node and edge costs. Given potential detections of humans and their parts separately, an initial set of pedestrian tracklets is first obtained, followed by explicit tracking of human parts as constrained by initial human tracking. A merging step is then performed whereby we attempt to include part-only detections for which the entire human is not observable. This step employs a selective appearance model, which allows us to skip occluded parts in description of positive training samples. The result is high confidence, robust trajectories of pedestrians as well as their parts, which essentially constrain each other's locations and associations, thus improving human tracking and parts detection. We test our algorithm on multiple real datasets and show that the proposed algorithm is an improvement over the state-of-the-art. © 2012 Springer-Verlag.

Publication Date

10-30-2012

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Volume

7577 LNCS

Issue

PART 6

Number of Pages

100-114

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.1007/978-3-642-33783-3_8

Socpus ID

84867880254 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/84867880254

This document is currently not available here.

Share

COinS