Online Multi-person Tracking by Tracker Hierarchy

Zhang, Jianming; Presti, Liliana Lo; Sclaroff, Stan

doi:10.1109/avss.2012.51

Cited by 63 publications

(61 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our novel occlusion-handling method improves the capability to deal with the occlusion of the tracker. Our occlusionhandling method makes the tracker more robust than the † http://www.cvg.rdg.ac.uk/PETS2009/ † † http://www.d2.mpi-inf.mpg.de/datasets [24] 71.3 67.5 100.0 0.0 method that employs a similar hierarchical data-association scheme [24]. Thus, we achieved a better MOTA score and a significantly improved MOTP score.…”

Section: Evaluation Metricsmentioning

confidence: 76%

“…The average speed is about 7 fps for all datasets. This performance also outperforms [13] (2 fps ), and on par with [4], [24] (7 fps). To the best of our knowledge, the major consuming time of our algorithm attributes to the DPM detector.…”

Section: Run Time Performancementioning

confidence: 89%

“…The classic online tracker [13] with a naive occlusion reasoning, a tracking framework employing similar hierarchical data association [24] without occlusion handling, the method utilizing scene knowledge [4] to tackle the occlusion issue, and the most recent approach [3]. As an example, a visual comparison between our method and Breitenstein et al [13] in two datasets illustrate the advantage of our proposed method.…”

Section: Evaluation Metricsmentioning

confidence: 99%

“…To the best of our knowledge, the major consuming time of our algorithm attributes to the DPM detector. On the other hand, [24] simply consider the hierarchical data association without considering the impact of social interaction, which significantly reduces the computation cost. Surprisingly, our performance could be on par with [4], which is offline method.…”

Section: Run Time Performancementioning

confidence: 99%

See 3 more Smart Citations

Inter-Person Occlusion Handling with Social Interaction for Online Multi-Pedestrian Tracking

Shen

2016

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARY Inter-person occlusion handling is a critical issue in the field of tracking, and it has been extensively researched. Several state-ofthe-art methods have been proposed, such as focusing on the appearance of the targets or utilizing knowledge of the scene. In contrast with the approaches proposed in the literature, we propose to address this issue using a social interaction model, which allows us to explore spatio-temporal information pertaining to the targets involved in the occlusion situation. Our experimental results show promising results compared with those obtained using other methods. key words : image processing, computer vision, inter-person occlusion, online tracking

show abstract

Section: Evaluation Metricsmentioning

confidence: 76%

Section: Run Time Performancementioning

confidence: 89%

Section: Evaluation Metricsmentioning

confidence: 99%

Section: Run Time Performancementioning

confidence: 99%

See 2 more Smart Citations

Inter-Person Occlusion Handling with Social Interaction for Online Multi-Pedestrian Tracking

Shen

2016

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

show abstract

“…To detect the ball and the book, we use region covariance templates [19] over the RGB-D channels. To deal with detection failures and maintain consistent labeling, we use the tracking-by-detection method proposed in [22], following depth-based background suppression. The tracker keeps an ensemble of appearance templates for each target and uses the tracker hierarchy to automatically handle tracker initialization, termination and tracking failure.…”

Section: Engagement Prediction In Ball and Book Playmentioning

confidence: 99%

Decoding Children's Social Behavior

Rehg

Abowd

Rozga

et al. 2013

2013 IEEE Conference on Computer Vision and Pattern Recognition

Self Cite

145

View full text Add to dashboard Cite

show abstract

Transfer learning‐based online multiperson tracking with Gaussian process regression

Zhang

Huang

et al. 2018

Concurrency and Computation

View full text Add to dashboard Cite

Summary Most existing tracking‐by‐detection approaches are affected by abrupt pedestrian pose changes, lighting conditions, scale changes, and real‐time processing, which leads to issues such as detection errors and drifts. To deal with these issues, we present a novel multi‐person tracking framework by introducing a new Gaussian Process Regression based observation model, which learns in a semi‐supervised manner. The background information is taken into consideration to build the discriminative tracker, training samples are re‐weighted appropriately to ease the impact of the potential sample misalignment and noisy during model updating. Unlabeled samples from the current frame provide rich information, which is used for enhancing the tracking inference. Experimental results show that the proposed approach outperforms a number of state‐of‐the‐art methods on some benchmark datasets.

show abstract

Online Multi-person Tracking by Tracker Hierarchy

Cited by 63 publications

References 19 publications

Inter-Person Occlusion Handling with Social Interaction for Online Multi-Pedestrian Tracking

Inter-Person Occlusion Handling with Social Interaction for Online Multi-Pedestrian Tracking

Decoding Children's Social Behavior

Transfer learning‐based online multiperson tracking with Gaussian process regression

Contact Info

Product

Resources

About