Staple: Complementary learners for real-time tracking
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 notoriou...
| Main Authors: | , , , , |
|---|---|
| Format: | Journal article |
| Published: |
IEEE
2016
|
| _version_ | 1797090031308898304 |
|---|---|
| author | Bertinetto, L Valmadre, J Golodetz, S Miksik, O Torr, P |
| author_facet | Bertinetto, L Valmadre, J Golodetz, S Miksik, O Torr, P |
| author_sort | Bertinetto, L |
| collection | OXFORD |
| description | 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. |
| first_indexed | 2024-03-07T03:12:37Z |
| format | Journal article |
| id | oxford-uuid:b4b1d03d-6a20-4fd4-be15-bee3788ae166 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T03:12:37Z |
| publishDate | 2016 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:b4b1d03d-6a20-4fd4-be15-bee3788ae1662022-03-27T04:28:11ZStaple: Complementary learners for real-time trackingJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:b4b1d03d-6a20-4fd4-be15-bee3788ae166Symplectic Elements at OxfordIEEE2016Bertinetto, LValmadre, JGolodetz, SMiksik, OTorr, PCorrelation 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. |
| spellingShingle | Bertinetto, L Valmadre, J Golodetz, S Miksik, O Torr, P Staple: Complementary learners for real-time tracking |
| title | Staple: Complementary learners for real-time tracking |
| title_full | Staple: Complementary learners for real-time tracking |
| title_fullStr | Staple: Complementary learners for real-time tracking |
| title_full_unstemmed | Staple: Complementary learners for real-time tracking |
| title_short | Staple: Complementary learners for real-time tracking |
| title_sort | staple complementary learners for real time tracking |
| work_keys_str_mv | AT bertinettol staplecomplementarylearnersforrealtimetracking AT valmadrej staplecomplementarylearnersforrealtimetracking AT golodetzs staplecomplementarylearnersforrealtimetracking AT miksiko staplecomplementarylearnersforrealtimetracking AT torrp staplecomplementarylearnersforrealtimetracking |