Abstract:One of the crucial problems in visual tracking is how the object is represented. Conventional appearance-based trackers are using increasingly more complex features in order to be robust. However, complex representations typically not only require more computation for feature extraction, but also make the state inference complicated. We show that with a careful feature selection scheme, extremely simple yet discriminative features can be used for robust object tracking. The central component of the proposed method is a succinct and discriminative representation of the object using discriminative non-orthogonal binary subspace (DNBS) which is spanned by Haar-like features. The DNBS representation inherits the merits of the original NBS in that it efficiently describes the object. It also incorporates the discriminative information to distinguish foreground from background. However, the problem of finding the DNBS bases from an over-complete dictionary is NP-hard. We propose a greedy algorithm called discriminative optimized orthogonal matching pursuit (D-OOMP) to solve this problem. An iterative formulation named iterative D-OOMP is further developed to drastically reduce the redundant computation between iterations and a hierarchical selection strategy is integrated for reducing the search space of features. The proposed DNBS representation is applied to object tracking through SSD-based template matching. We validate the effectiveness of our method through extensive experiments on challenging videos with comparisons against several state-of-the-art trackers and demonstrate its capability to track objects in clutter and moving background.

Robust Subspace Tracking Methods Based on Fast and Stable DPM Algorithm

A Novel Orthonormalization Matrix Based Fast and Stable DPM Algorithm for Principal and Minor Subspace Tracking

Robust visual tracking via weighted incremental subspace learning

Infrared Point Target Detection Based on Exponentially Weighted RLS Algorithm and Dual Solution Improvement

Impact of Data Channel on Tracking Performance of Pseudo-Code Tracking Loop

An Efficient Data-Driven Particle PHD Filter for Multitarget Tracking.

OPIT: A Simple but Effective Method for Sparse Subspace Tracking in High-Dimension and Low-Sample-Size Context

Low Complexity Robust Direction Finding Method for Impulsive Noise in Lp-Space

A New Look at the Power Method for Fast Subspace Tracking.

Tracking Randomly Moving Objects on Edge Box Proposals

Tracking in Multimedia Data Via Robust Reweighted Local Multi-Task Sparse Representation for Transportation Surveillance

A Robust SMC-PHD Filter for Multi-Target Tracking with Unknown Heavy-Tailed Measurement Noise

Seismic Noise Attenuation Using an Online Subspace Tracking Algorithm

Heavy-Tailed model for visual tracking via robust subspace learning

Robust Time Delay Estimation Based Direction Finding in L(P)-Space

MIMO Radar: An H ∞ Approach for Robust Multi-target Tracking in an Unknown Cluttered Environment

Multi-extended Targets PHD Filtering Tracking Algorithm with Unknown Measurement Noise Distribution

An Adaptive Tracking Method for Moving Target in Fluctuating Reverberation Environment

Infrared target tracking based on proximal robust principal component analysis method

Efficient Discriminative Nonorthogonal Binary Subspace with its Application to Visual Tracking

DOA Tracking Algorithm Based on AVS Pseudo-Smoothing for Coherent Acoustic Targets