Robust Structured Multi-Task Multi-View Sparse Tracking

Javanmardi, Mohammadreza; Qi, Xiaojun

doi:10.1109/icme.2018.8486581

Cited by 4 publications

(6 citation statements)

References 31 publications

(39 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This benchmark consists of 50 annotated sequences, where 49 sequences have one annotated target and one sequence (jogging) has two annotated targets [39]. We evaluate the overall performance of the proposed STLDF and its two variants (i.e., STLCF and STLHF) against 29 baseline trackers in Reference [39] and 17 recent trackers including MTMVTLS [43], MTMVTLAD [13], MSLA [14] (the recent version of ASLA [34]), SST [5], SMTMVT [44], CNT [21], two variants of TGPR (i.e., TGPR_Color and TGPR_HOG) [45], DSST [46], PCOM [47], KCF [19], MEEM [48], SAMF [49], SRDCF [50], STAPLE [51], and two variants of RSST (i.e., RSST_HOG and RSST_Deep) [23]. We present the overall OPE success and precision plots in Figure 2.…”

Section: Experimental Results On Otb50mentioning

confidence: 99%

A Robust Structured Tracker Using Local Deep Features

Javanmardi

Farzaneh

2020

Electronics

Self Cite

View full text Add to dashboard Cite

Deep features extracted from convolutional neural networks have been recently utilized in visual tracking to obtain a generic and semantic representation of target candidates. In this paper, we propose a robust structured tracker using local deep features (STLDF). This tracker exploits the deep features of local patches inside target candidates and sparsely represents them by a set of templates in the particle filter framework. The proposed STLDF utilizes a new optimization model, which employs a group-sparsity regularization term to adopt local and spatial information of the target candidates and attain the spatial layout structure among them. To solve the optimization model, we propose an efficient and fast numerical algorithm that consists of two subproblems with the close-form solutions. Different evaluations in terms of success and precision on the benchmarks of challenging image sequences (e.g., OTB50 and OTB100) demonstrate the superior performance of the STLDF against several state-of-the-art trackers.

show abstract

Section: Experimental Results On Otb50mentioning

confidence: 99%

A Robust Structured Tracker Using Local Deep Features

Javanmardi

Farzaneh

2020

Electronics

Self Cite

View full text Add to dashboard Cite

show abstract

“…We compare DF-SGLST with its two variants (SGLST Color and SGLST HOG), 29 baseline trackers in [42], and 17 recent trackers including MTMVTLS [18], MTMVTLAD [8], MSLA [20] (the recent version of ASLA [9]), SST [5], SMTMVT [46], CNT [34], two variants of TGPR (i.e., TGPR Color and TGPR HOG) [47], DSST [30], PCOM [48], KCF [28], MEEM [49], SAMF [50], SRDCF [51], STAPLE [52], and two variants of RSST (i.e., RSST HOG and RSST Deep) [10]. Following the protocol proposed in [42], we use the same parameters on all the sequences to obtain the one-pass evaluation (OPE) results, which are conventionally used to evaluate trackers by initializing them using the ground truth location in the first frame.…”

Section: Experimental Results On Otb50mentioning

confidence: 99%

“…For this benchmark data set, there are online available tracking results for 29 trackers [38]. In addition, we include the tracking results of additional 12 recent trackers, namely, MTMVTLS [31], MTMVTLAD [32], MSLA‐4 [35] (the recent version of ASLA [34]), SST [5], SMTMVT [47], CNT [48], TGPR [49], DSST [14], PCOM [50], KCF [45], MEEM [46], and RSST [36]. Following the protocol proposed in [38], we use the same parameters for SGLST_Color and SGLST_HOG on all the sequences to obtain the one‐pass evaluation (OPE) results, which are conventionally used to evaluate trackers by initialising them using the ground truth location in the first frame.…”

Section: Resultsmentioning

confidence: 99%

“…The proposed SGLST_HOG performs significantly better than traditional sparse trackers such as L1APG [29], LRST [30], ASLA [34], MTT [43], and MTMVTLS [31]. It outperforms most recent sparse trackers such as MTMVTLAD [32], SST [5], MSLA‐4 [35], SMTMVT [47], and RSST_HOG [36]. SGLST_HOG, which yields the AUC score of 0.556, also achieves better performance than some CF‐based methods such as KCF (AUC score of 0.514) and DSST (AUC score of 0.554).…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Structured group local sparse tracker

Javanmardi¹,

Qi²

2019

IET Image Processing

Self Cite

View full text Add to dashboard Cite

Sparse representation has recently been successfully applied in visual tracking. It utilizes a set of templates to represent target candidates and find the best one with the minimum reconstruction error as the tracking result. In this paper, we propose a robust deep features-based structured group local sparse tracker (DF-SGLST), which exploits the deep features of local patches inside target candidates and represents them by a set of templates in the particle filter framework. Unlike the conventional local sparse trackers, the proposed optimization model in DF-SGLST employs a group-sparsity regularization term to seamlessly adopt local and spatial information of the target candidates and attain the spatial layout structure among them. To solve the optimization model, we propose an efficient and fast numerical algorithm that consists of two subproblems with the closed-form solutions. Different evaluations in terms of success and precision on the benchmarks of challenging image sequences (e.g., OTB50 and OTB100) demonstrate the superior performance of the proposed tracker against several state-of-the-art trackers. Precision Precision plots of OPE -out of view (6) TGPR_HOG [0.785] MEEM [0.727] RSST_Deep [0.688] SRDCF [0.680] STAPLE [0.679] KCF [0.650] SMTMVT [0.642] SAMF [0.636] MTMVTLAD [0.607] RSST_HOG [0.600] DF-SGLST [0.588] TLD [0.576] MTMVTLS [0.573] LOT [0.567] Struck [0.539] LSK [0.515] DSST [0.511] CXT [0.510] TM-V [0.502] CNT [0.502] Fig. 3: The OPE success plots and precision plots for each of 11 challenge subsets in OTB50. Precision Precision plots of OPE -deformation (44) HDT [0.821] CNN-SVM [0.793] CF2 [0.791] MEEM [0.754] STAPLE [0.748] SRDCF [0.734] DF-SGLST [0.727] DLSSVM [0.727] RSST_Deep [0.717] LCT [0.689] SAMF [0.682] SGLST_HOG [0.628] TGPR_HOG [0.624] KCF [0.617] RSST_HOG [0.594] DSST [0.550] SGLST_Color [0.543] SCM [0.537] Struck [0.527] CPF [0.508]

show abstract

“…For example, it is used in head-pose classification [38,39], infer user's affective state [40], and video tracking problems [14,15]. And some recent works, such as [41,42,43] are also applied to the visual field.…”

Section: Multi-view Learningmentioning

confidence: 99%

Multi-view representation learning in multi-task scene

Liu

Lian

et al. 2019

Neural Comput & Applic

View full text Add to dashboard Cite

Over recent decades have witnessed considerable progress in whether multi-task learning or multiview learning, but the situation that consider both learning scenes simultaneously has received not too much attention. How to utilize multiple views' latent representation of each single task to improve each learning task's performance is a challenge problem. Based on this, we proposed a novel semi-supervised algorithm, termed as Multi-Task Multi-View learning based on Common and Special Features (MTMVCSF). In general, multi-views are the different aspects of an object and every view includes the underlying common or special information of this object. As a consequence, we will mine multiple views' jointly latent factor of each learning task which consists of each view's special feature and the common feature of all views. By this way, the original multi-task multi-view data has degenerated into multi-task data, and exploring the correlations among multiple tasks enables to make an improvement on the performance of learning algorithm. Another obvious advantage of this approach is that we get latent representation of the set of unlabeled instances by the constraint of regression task with labeled instances. The performance of classification and semi-supervised clustering task in these latent representations perform obviously better than it in raw data. Furthermore, an anti-noise multi-task multi-view algorithm called AN-MTMVCSF is proposed, which has a strong adaptability to noise labels. The effectiveness of these algorithms is proved by a series of well-designed experiments on both real world and synthetic data.

show abstract

Robust Structured Multi-Task Multi-View Sparse Tracking

Cited by 4 publications

References 31 publications

A Robust Structured Tracker Using Local Deep Features

A Robust Structured Tracker Using Local Deep Features

Structured group local sparse tracker

Multi-view representation learning in multi-task scene

Contact Info

Product

Resources

About