Real-Time Tracking via On-line Boosting
Published 1 January 2006
Helmut Gräbner, Michael Grabner, Horst Bischof
Citations1,208
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A novel on-line AdaBoost feature selection algorithm for tracking that allows to adapt the classifier while tracking the object and selects the most features for tracking resulting in stable tracking results.
Abstract
Very recently tracking was approached using classification techniques such as support vector machines. The object to be tracked is discriminated by a classifier from the background. In a similar spirit we propose a novel on-line AdaBoost feature selection algorithm for tracking. The distinct advantage of our method is its capability of on-line training. This allows to adapt the classifier while tracking the object. Therefore appearance changes of the object (e.g. out of plane rotations, illumination changes) are handled quite naturally. Moreover, depending on the background the algorithm selects the most discriminating features for tracking resulting in stable tracking results. By using fast computable features (e.g. Haar-like wavelets, orientation histograms, local binary patterns) the algorithm runs in real-time. We demonstrate the performance of the algorithm on several (publically available) video sequences. 1
Keywords
Computer Science
Histograms of Oriented Gradients for Human Detection
31,896 Citations2005Navneet Dalal, Bill Triggs
It is shown experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection, and the influence of each stage of the computation on performance is studied.
Rapid object detection using a boosted cascade of simple features
18,097 Citations2005Paul Viola, Michael Jones
A machine learning approach for visual object detection which is capable of processing images extremely rapidly and achieving high detection rates and the introduction of a new image representation called the "integral image" which allows the features used by the detector to be computed very quickly.
IEEE Transactions on Pattern Analysis and Machine IntelligenceMultiresolution gray-scale and rotation invariant texture classification with local binary patterns
15,254 Citations2002Timo Ojala, Matti Pietikäinen +1 more
A generalized gray-scale and rotation invariant operator presentation that allows for detecting the "uniform" patterns for any quantization of the angular space and for any spatial resolution and presents a method for combining multiple operators for multiresolution analysis.
Real-time tracking of non-rigid objects using mean shift
2,913 Citations2002Dorin Comaniciu, Visvanathan Ramesh +1 more
The theoretical analysis of the approach shows that it relates to the Bayesian framework while providing a practical, fast and efficient solution for real time tracking of non-rigid objects seen from a moving camera.
International Journal of Computer VisionDetecting Pedestrians Using Patterns of Motion and Appearance
2,104 Citations2005Paul Viola, Michael Jones +1 more
This pedestrian detection system is the first to combine both sources of information in a single detector, and operates on low resolution images under difficult conditions (such as rain and snow).
IEEE Transactions on Pattern Analysis and Machine IntelligenceOnline selection of discriminative tracking features
1,217 Citations2005Robert T. Collins, Yanxi Liu +1 more
This paper presents an online feature selection mechanism for evaluating multiple features while tracking and adjusting the set of features used to improve tracking performance, and notes susceptibility of the variance ratio feature selection method to distraction by spatially correlated background clutter.
IEEE Transactions on Pattern Analysis and Machine IntelligenceEfficient region tracking with parametric models of geometry and illumination
1,061 Citations1998Gregory D. Hager, Peter N. Belhumeur
This work develops a computationally efficient method for handling the geometric distortions produced by changes in pose and combines geometry and illumination into an algorithm that tracks large image regions using no more computation than would be required to track with no accommodation for illumination changes.
IEEE Transactions on Pattern Analysis and Machine IntelligenceSupport vector tracking
1,048 Citations2004Shai Avidan
Support Vector Tracking integrates the Support Vector Machine (SVM) classifier into an optic-flow-based tracker and maximizes the SVM classification score to account for large motions between successive frames.
International Journal of Computer VisionEigenTracking: Robust Matching and Tracking of Articulated Objects Using a View-Based Representation
1,032 Citations1998Michael J. Black, Allan D. Jepson
A “subspace constancy assumption” is defined that allows techniques for parameterized optical flow estimation to simultaneously solve for the view of an object and the affine transformation between the eigenspace and the image.
Mean shift analysis and applications
986 Citations1999Dorin Comaniciu, Peter Meer
A nonparametric estimator of density gradient, the mean shift, is employed in the joint, spatial-range (value) domain of gray level and color images for discontinuity preserving filtering and image segmentation and its convergence on lattices is proven.
On-line Boosting and Vision
984 Citations2006Helmut Gräbner, Horst Bischof
This paper proposes a novel on-line AdaBoost feature selection method and demonstrates the multifariousness of the method on such diverse tasks as learning complex background models, visual tracking and object detection.
IEEE Transactions on Pattern Analysis and Machine IntelligenceThe template update problem
762 Citations2004Linda Matthews, Takahiro Ishikawa +1 more
Integral histogram: a fast way to extract histograms in Cartesian spaces
701 Citations2005Fatih Porikli
The integral histogram method makes it possible to employ even an exhaustive search process in real-time, which was impractical before, and enables the description of higher level histogram features.
Learning object detection from a small number of examples: the importance of good features
380 Citations2004K. Levi, Yaakov Weiss
This work shows that using local edge orientation histograms (EOH) as features can significantly improve performance compared to the standard linear features used in existing systems and enables learning a system that seems to outperform the state of the art in real-time systems even with a small number of training examples.
Ensemble Tracking
320 Citations2005Shai Avidan
This work considers tracking as a binary classification problem, where an ensemble of weak classifiers is trained online to distinguish between the object and the background, and combines them into a strong classifier using AdaBoost.
Incremental Learning for Visual Tracking
287 Citations2004Jongwoo Lim, David A. Ross +2 more
This paper presents an efficient and effective online algorithm that incrementally learns and adapts a low dimensional eigenspace representation to reflect appearance changes of the target, thereby facilitating the tracking task.
Robust online appearance models for visual tracking
248 Citations2005Allan D. Jepson, David J. Fleet +1 more
A framework for learning robust, adaptive appearance models to be used for motion-based tracking of natural objects and provides the ability to adapt to natural changes in appearance, such as those due to facial expressions or variations in 3D pose.
IEEE Transactions on Pattern Analysis and Machine IntelligenceSparse Bayesian learning for efficient visual tracking
195 Citations2005Oliver Williams, Andrew Blake +1 more
This paper builds a displacement expert which directly estimates displacement from the target region and is demonstrated in real-time tracking systems where the sparsity of the RVM means that only a fraction of CPU time is required to track at frame rate.
Probabilistic tracking in joint feature-spatial spaces
171 Citations2003Ahmed Elgammal, Ramani Duraiswami +1 more
This paper exploits the feature-spatial distribution of a region representing an object as a probabilistic constraint to track that region over time and applies it to tracking with different types of image features including row intensity, color and image gradient.
Visual tracking using learned linear subspaces
166 Citations2004J. Ho, Kuang-Chih Lee +2 more
This paper presents a simple but robust visual tracking algorithm based on representing the appearances of objects using affine warps of learned linear subspaces of the image space, and argues that a variant of it, the uniform L/sup 2/-reconstruction error norm, is the right one for tracking.
Learning a Rare Event Detection Cascade by Direct Feature Selection
126 Citations2003Jianxin Wu, James M. Rehg +1 more
A novel cascade learning algorithm based on forward feature selection which is two orders of magnitude faster than the Viola-Jones approach and yields classifiers of equivalent quality could be used for more demanding classification tasks, such as on-line learning.
Online Selecting Discriminative Tracking Features Using Particle Filter
119 Citations2005Jianyu Wang, Xilin Chen +1 more
The paper proposes a method to keep the tracker robust to background clutters by online selecting discriminative features from a large feature space by embedded into the particle filtering process with the aid of existed "background" particles.
On-line density-based appearance modeling for object tracking
102 Citations2005Bohyung Han, L.S. Davis
This work presents a new on-line appearance modeling technique which is based on sequential density approximation, which provides accurate and compact representations using Gaussian mixtures, in which the number of Gaussians is automatically determined.
Lecture notes in computer scienceTracking Aspects of the Foreground against the Background
43 Citations2004Hieu T. Nguyen, A.W.M. Smeulders
The paper proposes an approach to this problem without a priori learning object views that is robust to even severe aspect changes when unseen views of the object come into view.