Visual Tracking Via Kernel Sparse Representation With Multikernel Fusion

Wang, Lingfeng; Yan, Hongping; Lv, Ke; Chen, Pan

doi:10.1109/tcsvt.2014.2302496

Cited by 31 publications

(14 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…x t is convolved with its corresponding filter. Then, the multi-channel convolution score   f Sx is obtained by summing the result of all channels, as defined in (2). Therefore, the correlation filtering optimal function is the proposed formulation as follows:…”

Section: A Cf Trackersmentioning

confidence: 99%

See 1 more Smart Citation

Adaptive and Complementary Correlation Filter With Dynamic Contextual Constraints

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Correlation filter-based trackers have gained more and more attention because of their great performance and relative high tracking speed. However, this kind of trackers may suffer model drifting due to learning limited background information during filter training. This may lead to tracking failures in some complex scenes, such as background clutter, deformation, illumination variation and so on. In this paper, we propose an adaptive and complementary correlation filter with dynamic contextual constraints. First, we introduce contextual information around the target as a dynamic constrained term to alleviate model drifting in complex scenes, the optimal function of which can be solved by iterative method. Then, we integrate a color histogram-based tracker to compensate the inaccurate tracking of correlation filtering. In addition, we present metrics to combine the two complementary trackers with adaptive fusion coefficients. Finally, extensive experiments on OTB2013, OTB2015, VOT2016 and UAV123 benchmark datasets demonstrate that our tracker can improve the performance of our baseline and can perform favorably against some stateof-the-art trackers.

show abstract

Section: A Cf Trackersmentioning

confidence: 99%

“…There are various different tracking algorithms. Like generative methods that construct target templates and search for the most similar patches as tracking result [1], [2]. Moreover, there are Siamese network based trackers that have end-to-end training capabilities and high efficiency [3], [4].…”

Section: Introductionmentioning

confidence: 99%

Adaptive and Complementary Correlation Filter With Dynamic Contextual Constraints

et al. 2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…3) It cannot achieve superior results for fast moving object. 4) It is hard to achieve real -time speed [9]. Longyin Wen et al proposed a robust spatio -temporal context model based tracker to complete the tracking task in unconstrained environments.…”

Section: Review Of Literature Surveymentioning

confidence: 99%

Tracking Moving Objects in Video

Jemilda¹

2017

JCP

View full text Add to dashboard Cite

Object tracking is a well-studied problem in computer vision and has many practical applications. The problem and its difficulty depend upon several factors such as the knowledge about the target object, its quantity and type of parameters being tracked. Although there has been some success with building trackers for specific object classes such as human, face, mice etc. Tracking generic objects has remained challenging issue because an object can drastically change its appearance when deforming, rotating out of plane or when the illumination of the scene changes. Especially in the videos which are not clear in its original form i.e. the video which requires enhancement and its quality has to be improved. Here in the proposed work different algorithms are carried out to detect the generic objects such as the non-rigid objects and the deformable objects which are in occlusion, (i.e. in a cluttered environment) not only in the images which contain some sort of objects but also from the images which contains numerous objects which are unseen so as to improve the tracking efficiency. By doing this, the objects which are non-rigid and changing in nature can also be predicted in a perfect manner so that the computational complexity can be reduced, reliability, accuracy, and efficiency can be improved.

show abstract

“…Inspired by the success of sparse representation in face recognition [33], recently, sparse representation based visual tracking becomes overwhelming [28], [31], [32], [34]- [36]. The first sparse representation based tracking method was presented in [28], which is implemented under the widely used particle filter framework [37], [38] and represents each target candidate (corresponding to a particle) y using a set of target and trivial templates.…”

Section: Related Workmentioning

confidence: 99%

Robust Visual Tracking Using Structurally Random Projection and Weighted Least Squares

Zhang

Zhou

Jiang

et al. 2015

IEEE Trans. Circuits Syst. Video Technol.

122

View full text Add to dashboard Cite

Sparse representation based visual tracking approaches have attracted increasing interests in the community in recent years. The main idea is to linearly represent each target candidate using a set of target and trivial templates while imposing a sparsity constraint onto the representation coefficients. After we obtain the coefficients using L1-norm minimization methods, the candidate with the lowest error, when it is reconstructed using only the target templates and the associated coefficients, is considered as the tracking result. In spite of promising system performance widely reported, it is unclear if the performance of these trackers can be maximised. In addition, computational complexity caused by the dimensionality of the feature space limits these algorithms in real-time applications. In this paper, we propose a real-time visual tracking method based on structurally random projection and weighted least squares techniques. In particular, to enhance the discriminative capability of the tracker, we introduce background templates to the linear representation framework. To handle appearance variations over time, we relax the sparsity constraint using a weighed least squares (WLS) method to obtain the representation coefficients. To further reduce the computational complexity, structurally random projection is used to reduce the dimensionality of the feature space while preserving the pairwise distances between the data points in the feature space. Experimental results show that the proposed approach outperforms several state-of-the-art tracking methods.

show abstract

Visual Tracking Via Kernel Sparse Representation With Multikernel Fusion

Cited by 31 publications

References 22 publications

Adaptive and Complementary Correlation Filter With Dynamic Contextual Constraints

Adaptive and Complementary Correlation Filter With Dynamic Contextual Constraints

Tracking Moving Objects in Video

Robust Visual Tracking Using Structurally Random Projection and Weighted Least Squares

Contact Info

Product

Resources

About