Adaptive Probabilistic Visual Tracking with Incremental Subspace Update

David Ross, Jongwoo Lim, Ming Hsuan Yang

Research output: Contribution to journalArticle

84 Citations (Scopus)

Abstract

Visual tracking, in essence, deals with non-stationary data streams that change over time. While most existing algorithms are able to track objects well in controlled environments, they usually fail if there is a significant change in object appearance or surrounding illumination. The reason being that these visual tracking algorithms operate on the premise that the models of the objects being tracked are invariant to internal appearance change or external variation such as lighting or viewpoint. Consequently most tracking algorithms do not update the models once they are built or learned at the outset. In this paper, we present an adaptive probabilistic tracking algorithm that updates the models using an incremental update of eigenbasis. To track objects in two views, we use an effective probabilistic method for sampling affine motion parameters with priors and predicting its location with a maximum a posteriori estimate. Borne out by experiments, we demonstrate the proposed method is able to track objects well under large lighting, pose and scale variation with close to real-time performance.

Original languageEnglish
Pages (from-to)470-482
Number of pages13
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3022
Publication statusPublished - 2004 Dec 1

Fingerprint

Visual Tracking
Update
Subspace
Lighting
A Posteriori Estimates
Maximum a Posteriori
Probabilistic Methods
Data Streams
Sampling
Illumination
Object
Model
Internal
Real-time
Invariant
Motion
Experiments
Demonstrate
Experiment

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

@article{e784c3d336f941b1b7fd1c8c552d5bff,
title = "Adaptive Probabilistic Visual Tracking with Incremental Subspace Update",
abstract = "Visual tracking, in essence, deals with non-stationary data streams that change over time. While most existing algorithms are able to track objects well in controlled environments, they usually fail if there is a significant change in object appearance or surrounding illumination. The reason being that these visual tracking algorithms operate on the premise that the models of the objects being tracked are invariant to internal appearance change or external variation such as lighting or viewpoint. Consequently most tracking algorithms do not update the models once they are built or learned at the outset. In this paper, we present an adaptive probabilistic tracking algorithm that updates the models using an incremental update of eigenbasis. To track objects in two views, we use an effective probabilistic method for sampling affine motion parameters with priors and predicting its location with a maximum a posteriori estimate. Borne out by experiments, we demonstrate the proposed method is able to track objects well under large lighting, pose and scale variation with close to real-time performance.",
author = "David Ross and Jongwoo Lim and Yang, {Ming Hsuan}",
year = "2004",
month = "12",
day = "1",
language = "English",
volume = "3022",
pages = "470--482",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Adaptive Probabilistic Visual Tracking with Incremental Subspace Update

AU - Ross, David

AU - Lim, Jongwoo

AU - Yang, Ming Hsuan

PY - 2004/12/1

Y1 - 2004/12/1

N2 - Visual tracking, in essence, deals with non-stationary data streams that change over time. While most existing algorithms are able to track objects well in controlled environments, they usually fail if there is a significant change in object appearance or surrounding illumination. The reason being that these visual tracking algorithms operate on the premise that the models of the objects being tracked are invariant to internal appearance change or external variation such as lighting or viewpoint. Consequently most tracking algorithms do not update the models once they are built or learned at the outset. In this paper, we present an adaptive probabilistic tracking algorithm that updates the models using an incremental update of eigenbasis. To track objects in two views, we use an effective probabilistic method for sampling affine motion parameters with priors and predicting its location with a maximum a posteriori estimate. Borne out by experiments, we demonstrate the proposed method is able to track objects well under large lighting, pose and scale variation with close to real-time performance.

AB - Visual tracking, in essence, deals with non-stationary data streams that change over time. While most existing algorithms are able to track objects well in controlled environments, they usually fail if there is a significant change in object appearance or surrounding illumination. The reason being that these visual tracking algorithms operate on the premise that the models of the objects being tracked are invariant to internal appearance change or external variation such as lighting or viewpoint. Consequently most tracking algorithms do not update the models once they are built or learned at the outset. In this paper, we present an adaptive probabilistic tracking algorithm that updates the models using an incremental update of eigenbasis. To track objects in two views, we use an effective probabilistic method for sampling affine motion parameters with priors and predicting its location with a maximum a posteriori estimate. Borne out by experiments, we demonstrate the proposed method is able to track objects well under large lighting, pose and scale variation with close to real-time performance.

UR - http://www.scopus.com/inward/record.url?scp=35048823490&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35048823490&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:35048823490

VL - 3022

SP - 470

EP - 482

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -