Visual Object Tracking with Online Learning on Riemannian Manifolds by One-Class Support Vector Machines
Paper i proceeding, 2014

This paper addresses issues in video object tracking. We propose a novel method where tracking is regarded as a one-class classification problem of domain-shift objects. The proposed tracker is inspired by the fact that the positive samples can be bounded by a closed hypersphere generated by one-class support vector machines (SVM), leading to a solution for robust learning of target model online. The main novelties of the paper include: (a) represent the target model by a set of positive samples as a cluster of points on Riemannian manifolds; (b) perform online learning of target model as a dynamic cluster of points flowing on the manifold, in an alternate manner with tracking; (c) formulate geodesic-based kernel function for one-class SVM on Riemannian manifolds under the log-Euclidean metric. Experiments are conducted on several videos, results have provided support to the proposed method.

Riemannian manifold

support vector machines

online learning

one-class classification

covariance matrix

Visual object tracking

Författare

Yixiao Yun

Chalmers, Signaler och system, Signalbehandling och medicinsk teknik

Keren Fu

Chalmers, Signaler och system, Signalbehandling och medicinsk teknik

Irene Yu-Hua Gu

Chalmers, Signaler och system, Signalbehandling och medicinsk teknik

Jie Yang

Shanghai Jiao Tong University

IEEE International Conference on Image Processing (ICIP 2014), Oct.27 - 30, 2014, Paris, France

1902-1906
978-147995751-4 (ISBN)

Styrkeområden

Informations- och kommunikationsteknik

Transport

Ämneskategorier

Geometri

Systemvetenskap

Signalbehandling

Datorseende och robotik (autonoma system)

DOI

10.1109/ICIP.2014.7025381

ISBN

978-147995751-4

Mer information

Senast uppdaterat

2018-03-07