TRIC-track: tracking by regression with incrementally learned cascades

This paper proposes a novel approach to part-based track- ing by replacing local matching of an appearance model by direct prediction of the displacement between local image patches and part locations. We propose to use cascaded regression with incremental learning to track generic objects without a...

Full description

Bibliographic Details
Main Authors: Wang, Xiaomeng, Valstar, Michel F., Martinez, Brais, Khan, Muhammad Haris, Pridmore, Tony
Format: Conference or Workshop Item
Published: 2015
Online Access:https://eprints.nottingham.ac.uk/31307/
_version_ 1848794173320200192
author Wang, Xiaomeng
Valstar, Michel F.
Martinez, Brais
Khan, Muhammad Haris
Pridmore, Tony
author_facet Wang, Xiaomeng
Valstar, Michel F.
Martinez, Brais
Khan, Muhammad Haris
Pridmore, Tony
author_sort Wang, Xiaomeng
building Nottingham Research Data Repository
collection Online Access
description This paper proposes a novel approach to part-based track- ing by replacing local matching of an appearance model by direct prediction of the displacement between local image patches and part locations. We propose to use cascaded regression with incremental learning to track generic objects without any prior knowledge of an object’s structure or appearance. We exploit the spatial constraints between parts by implicitly learning the shape and deformation parameters of the object in an online fashion. We integrate a multiple temporal scale motion model to initialise our cascaded regression search close to the target and to allow it to cope with occlusions. Experimental results show that our tracker ranks first on the CVPR 2013 Benchmark.
first_indexed 2025-11-14T19:11:59Z
format Conference or Workshop Item
id nottingham-31307
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:11:59Z
publishDate 2015
recordtype eprints
repository_type Digital Repository
spelling nottingham-313072020-05-04T20:06:19Z https://eprints.nottingham.ac.uk/31307/ TRIC-track: tracking by regression with incrementally learned cascades Wang, Xiaomeng Valstar, Michel F. Martinez, Brais Khan, Muhammad Haris Pridmore, Tony This paper proposes a novel approach to part-based track- ing by replacing local matching of an appearance model by direct prediction of the displacement between local image patches and part locations. We propose to use cascaded regression with incremental learning to track generic objects without any prior knowledge of an object’s structure or appearance. We exploit the spatial constraints between parts by implicitly learning the shape and deformation parameters of the object in an online fashion. We integrate a multiple temporal scale motion model to initialise our cascaded regression search close to the target and to allow it to cope with occlusions. Experimental results show that our tracker ranks first on the CVPR 2013 Benchmark. 2015-12 Conference or Workshop Item PeerReviewed Wang, Xiaomeng, Valstar, Michel F., Martinez, Brais, Khan, Muhammad Haris and Pridmore, Tony (2015) TRIC-track: tracking by regression with incrementally learned cascades. In: International Conference on Computer Vision (ICCV15), 11-18 December 2015, Santiago, Chile. http://www.cv-foundation.org/openaccess/content_iccv_2015/papers/Wang_TRIC-track_Tracking_by_ICCV_2015_paper.pdf
spellingShingle Wang, Xiaomeng
Valstar, Michel F.
Martinez, Brais
Khan, Muhammad Haris
Pridmore, Tony
TRIC-track: tracking by regression with incrementally learned cascades
title TRIC-track: tracking by regression with incrementally learned cascades
title_full TRIC-track: tracking by regression with incrementally learned cascades
title_fullStr TRIC-track: tracking by regression with incrementally learned cascades
title_full_unstemmed TRIC-track: tracking by regression with incrementally learned cascades
title_short TRIC-track: tracking by regression with incrementally learned cascades
title_sort tric-track: tracking by regression with incrementally learned cascades
url https://eprints.nottingham.ac.uk/31307/
https://eprints.nottingham.ac.uk/31307/