Tracking People and Their Objects

Baumgärtner, Tobias; Mitzel, Dennis; Leibe, Bastian

doi:10.1109/cvpr.2013.469

Cited by 13 publications

(9 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chang M. et al [6] address the challenge of recognising the behaviour of groups of individuals in surveillance environments proposing a novel grouping metric and using track-based motion analysis. There is also other recent work by researchers dealing with the problem of people tracking [7,8,9]. Often, people tracking by using a single camera is insufficient for understanding people's behaviour in an environment and a set of cameras is used to deal with this problem.…”

Section: A Video Analytic Toolsmentioning

confidence: 99%

A rule-based methodology and assessment for context-aware privacy

Teixeira

Maffra

Lelu

et al. 2014

2014 IEEE 6th International Conference on Awareness Science and Technology (iCAST)

View full text Add to dashboard Cite

High resolution surveillance systems are essential for security. However, these powerful tools have been misused by several CCTV operators. The governments and civil society are attempting to strike a balance between safety and privacy. Privacy filters can be used to help protect part of an image which included Personally Identifiable Information (PII). This paper presents a novel approach to improve the privacy protection in the CCTV displays. Our method uses context cues to determine the required privacy filtering level for each person in the image. We also present a systematic methodology to handle the context cues. We use a rules engine to generalise and facilitate the customisation of this system that by design should be specialized for operation in a given environment. In addition, we present a case study as a proof-of-concept whereby we have created an environment providing with high levels of privacy protection whilst allowing the required level of surveillance monitoring.

show abstract

Section: A Video Analytic Toolsmentioning

confidence: 99%

A rule-based methodology and assessment for context-aware privacy

Teixeira

Maffra

Lelu

et al. 2014

2014 IEEE 6th International Conference on Awareness Science and Technology (iCAST)

View full text Add to dashboard Cite

show abstract

“…The approach of [9] exploits person-to-person and personto-object interactions to more reliably track both people and objects. It relies on a Bayesian Network model to enforce frame-to-frame temporal consistency, and on training data to learn object types and appearances.…”

Section: Tracking Interacting Objectsmentioning

confidence: 99%

“…This is much more general than what is done in approaches such as [7], in which the appearance of people is used to infer the possible presence of a static entrance. It also goes beyond recent work on interaction between people and objects [9]. Due to the global nature of the optimization and the generality of the constraints, we can deal with objects that may be completely hidden during large portions of the interaction and do not require any training data.…”

Section: Introductionmentioning

confidence: 99%

Tracking Interacting Objects Using Intertwined Flows

Wang

Türetken

Fleuret

et al. 2016

IEEE Trans. Pattern Anal. Mach. Intell.

127

View full text Add to dashboard Cite

Abstract-In this paper, we show that tracking different kinds of interacting objects can be formulated as a network-flow Mixed Integer Program. This is made possible by tracking all objects simultaneously using intertwined flow variables and expressing the fact that one object can appear or disappear at locations where another is in terms of linear flow constraints. Our proposed method is able to track invisible objects whose only evidence is the presence of other objects that contain them. Furthermore, our tracklet-based implementation yields real-time tracking performance. We demonstrate the power of our approach on scenes involving cars and pedestrians, bags being carried and dropped by people, and balls being passed from one player to the next in team sports. In particular, we show that by estimating jointly and globally the trajectories of different types of objects, the presence of the ones which were not initially detected based solely on image evidence can be inferred from the detections of the others.

show abstract

“…However, this is very different from having objects that may move, thereby allowing objects of a different nature to appear or disappear at varying locations. In [2], person-to-person and person-to-object interactions are exploited to more reliably track all of them. This approach relies on a Bayesian Network model to enforce frame-to-frame temporal coherence, and on training data to learn object types and appearances.…”

Section: Related Workmentioning

confidence: 99%

“…This is much more general than what is done in approaches such as [28], in which the appearance of people is used to infer the possible presence of a static entrance. It also goes beyond recent work on interaction between people and objects [2]. Due to the global nature of the optimization and the generality of the constraints, we can deal with objects that may be completely hidden during large portions of the interaction and do not require any training data.…”

Section: Introductionmentioning

confidence: 99%

Tracking Interacting Objects Optimally Using Integer Programming

Wang

Türetken

Fleuret

et al. 2014

Computer Vision – ECCV 2014

View full text Add to dashboard Cite

Abstract. In this paper, we show that tracking different kinds of interacting objects can be formulated as a network-flow Mixed Integer Program. This is made possible by tracking all objects simultaneously and expressing the fact that one object can appear or disappear at locations where another is in terms of linear flow constraints. We demonstrate the power of our approach on scenes involving cars and pedestrians, bags being carried and dropped by people, and balls being passed from one player to the next in a basketball game. In particular, we show that by estimating jointly and globally the trajectories of different types of objects, the presence of the ones which were not initially detected based solely on image evidence can be inferred from the detections of the others.

show abstract

Tracking People and Their Objects

Cited by 13 publications

References 19 publications

A rule-based methodology and assessment for context-aware privacy

A rule-based methodology and assessment for context-aware privacy

Tracking Interacting Objects Using Intertwined Flows

Tracking Interacting Objects Optimally Using Integer Programming

Contact Info

Product

Resources

About