AdaPT: Real-time adaptive pedestrian tracking for crowded scenes

Bera, Aniket; Galoppo, Nico; Sharlet, Dillon; Lake, Adam; Manocha, Dinesh

doi:10.1109/icra.2014.6907095

Cited by 34 publications

(23 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our appearance term is based on an online discriminative learning approach, where we train a regression model for each target. This is different from previous works, where a generative model such as template based tracking is used as a baseline [11], [12], [13], [14]. The second and third components in our objective function capture the motion of the crowd.…”

Section: Arxiv:160309240v1 [Cscv] 30 Mar 2016mentioning

confidence: 93%

“…Additionally, the large number of targets to be tracked increases the computational complexity and makes the design of an efficient tracker even more challenging. The latter issue, is one of the main reasons that all previous trackers [11], [12], [13], [14], focused on high-density crowds, track one target at a time instead of jointly optimizing the objective function for all the targets simultaneously. Although focusing on tracking one target at a time helps reduce the computational complexity, joint optimization is essential for optimal multi-target tracking.…”

Section: Arxiv:160309240v1 [Cscv] 30 Mar 2016mentioning

confidence: 99%

“…For example, multiple targets cannot occupy the same location at the same time. Thus for optimal assignments of new locations, objects in the scene should compete for each candidate location simultaneously, which is not the case in [11], [12], [13], [14]. Additionally modeling interactions between targets can greatly benefit multi-target tracking algorithms.…”

Section: Arxiv:160309240v1 [Cscv] 30 Mar 2016mentioning

confidence: 99%

“…Target tracking in highly crowded scene is relatively a new area of research, and only a handful of papers have focused on this problem [11], [12], [13], [14]. The methods proposed in [11], [12], [13], [14] track each target separately by training an online tracker for each individual separately.…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Binary Quadratic Programing for Online Tracking of Hundreds of People in Extremely Crowded Scenes

Dehghan

Shah

2018

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Abstract-Multi-object tracking has been studied for decades. However, when it comes to tracking pedestrians in extremely crowded scenes, we are limited to only few works. This is an important problem which gives rise to several challenges. Pre-trained object detectors fail to localize targets in crowded sequences. This consequently limits the use of data-association based multi-target tracking methods which rely on the outcome of an object detector. Additionally, the small apparent target size makes it challenging to extract features to discriminate targets from their surroundings. Finally, the large number of targets greatly increases computational complexity which in turn makes it hard to extend existing multi-target tracking approaches to high-density crowd scenarios. In this paper, we propose a tracker that addresses the aforementioned problems and is capable of tracking hundreds of people efficiently. We formulate online crowd tracking as Binary Quadratic Programing. Our formulation employs target's individual information in the form of appearance and motion as well as contextual cues in the form of neighborhood motion, spatial proximity and grouping constraints, and solves detection and data association simultaneously. In order to solve the proposed quadratic optimization efficiently, where state-of art commercial quadratic programing solvers fail to find the answer in a reasonable amount of time, we propose to use the most recent version of the Modified Frank Wolfe algorithm, which takes advantage of SWAP-steps to speed up the optimization. We show that the proposed formulation can track hundreds of targets efficiently and improves state-of-art results by significant margins on eleven challenging high density crowd sequences.

show abstract

Section: Arxiv:160309240v1 [Cscv] 30 Mar 2016mentioning

confidence: 93%

Section: Arxiv:160309240v1 [Cscv] 30 Mar 2016mentioning

confidence: 99%

Section: Arxiv:160309240v1 [Cscv] 30 Mar 2016mentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Binary Quadratic Programing for Online Tracking of Hundreds of People in Extremely Crowded Scenes

Dehghan

Shah

2018

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…Alternatively, a hybrid detector may be built by switching entirely among more available detectors based on the current estimated difficulty of the scene [9]. In the context of online multi-target tracking, [10] proposes to exploit a batch of classifiers and obtain strong and weak detections on the basis of a confidence score.…”

Section: B Fusionmentioning

confidence: 99%