We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Staple: Complementary Learners for Real-Time Tracking

Abstract: Correlation Filter-based trackers have recently achieved excellent performance, showing great robustness to challenging situations exhibiting motion blur and illumination changes. However, since the model that they learn depends strongly on the spatial layout of the tracked object, they are notoriously sensitive to deformation. Models based on colour statistics have complementary traits: they cope well with variation in shape, but suffer when illumination is not consistent throughout a sequence. Moreover, colour distributions alone can be insufficiently discriminative. In this paper, we show that a simple tracker combining complementary cues in a ridge regression framework can operate faster than 80 FPS and outperform not only all entries in the popular VOT14 competition, but also recent and far more sophisticated trackers according to multiple benchmarks.
Comments: To appear in CVPR 2016
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:1512.01355 [cs.CV]
  (or arXiv:1512.01355v2 [cs.CV] for this version)

Submission history

From: Luca Bertinetto [view email]
[v1] Fri, 4 Dec 2015 09:56:48 GMT (769kb,D)
[v2] Wed, 13 Apr 2016 17:58:11 GMT (1487kb,D)

Link back to: arXiv, form interface, contact.