Online Bayesian Non-parametrics for Social Group Detection

Zanotto, Matteo; Bazzani, Loris; Cristani, Marco; Murino, Vittorio

doi:10.5244/c.26.111

Cited by 19 publications

(29 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In practice, this variant separates individual tracking from group tracking in two different particle filters. Moreover, we reported the results using the DPMM detector [34] applied to the individual tracks given by the DEEPER-JIGT (last row in Table 1). …”

Section: Results: Synthetic Scenariosmentioning

confidence: 99%

“…This data association-based tracker uses the detections of groups given by the DPMM detector [34]. Group tracking is performed by associating the groups at time t with the groups at time t − 1 through nearest neighbor on the position-velocity vector.…”

Section: Results: Real Scenariosmentioning

confidence: 99%

“…To pursue this, we use the sequential variational inference algorithm presented in [34] which allows us to implement online unsupervised learning of group structures. The algorithm, based on [60], relies on a truncation of the stick-breaking construction.…”

Section: Online Inferencementioning

confidence: 99%

“…The second one fuses the DPF with a Dirichlet Process Mixture Model (DPMM) [34], that learns in an online fashion the group configuration without the need for any pre-trained classifier. This differentiation gives rise to two different Joint IndividualGroup Trackers (JIGT), called DEcentralizEd Particle filtER (DEEPER-JIGT) and Dirichlet Process-based Decentralized Particle filter (DP2-JIGT).…”

mentioning

confidence: 99%

See 3 more Smart Citations

Joint Individual-Group Modeling for Tracking

Bazzani

Zanotto

Cristani

et al. 2015

IEEE Trans. Pattern Anal. Mach. Intell.

Self Cite

View full text Add to dashboard Cite

Abstract-We present a novel probabilistic framework that jointly models individuals and groups for tracking. Managing groups is challenging, primarily because of their nonlinear dynamics and complex layout which lead to repeated splitting and merging events. The proposed approach assumes a tight relation of mutual support between the modeling of individuals and groups, promoting the idea that groups are better modeled if individuals are considered and vice versa. This concept is translated in a mathematical model using a decentralized particle filtering framework which deals with a joint individual-group state space. The model factorizes the joint space into two dependent subspaces, where individuals and groups share the knowledge of the joint individual-group distribution. The assignment of people to the different groups (and thus group initialization, split and merge) is implemented by two alternative strategies: using classifiers trained beforehand on statistics of group configurations, and through online learning of a Dirichlet process mixture model, assuming that no training data is available before tracking. These strategies lead to two different methods that can be used on top of any person detector (simulated using the ground truth in our experiments). We provide convincing results on two recent challenging tracking benchmarks.

show abstract

Section: Results: Synthetic Scenariosmentioning

confidence: 99%

Section: Results: Real Scenariosmentioning

confidence: 99%

Section: Online Inferencementioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Joint Individual-Group Modeling for Tracking

Bazzani

Zanotto

Cristani

et al. 2015

IEEE Trans. Pattern Anal. Mach. Intell.

Self Cite

View full text Add to dashboard Cite

show abstract

“…A modified agglomerative clustering approach is then performed to infer pedestrian groups. Zanotto et al [25] introduced an unsupervised approach based on an online inference process of Dirichlet Process Mixture Models.…”

Section: Introductionmentioning

confidence: 99%

Social Group Discovery from Surveillance Videos: A Data-Driven Approach with Attention-Based Cues

Chamveha¹,

Sugano²,

Sato³

et al. 2013

Procedings of the British Machine Vision Conference 2013

View full text Add to dashboard Cite

This paper presents an approach to discover social groups in surveillance videos by incorporating attention-based cues to model group behaviors of pedestrians in videos. Group behaviors are modeled as a set of decision trees with the decisions being basic measurements based on position-based and attention-based cues. Rather than enforcing explicit models, we apply tree-based learning algorithms to implicitly construct the decision tree models. The experimental results demonstrate that incorporating attentionbased cues significantly increased the estimation accuracy compared to the conventional approaches that used position-based cues alone.

show abstract

GD-GAN: Generative Adversarial Networks for Trajectory Prediction and Group Detection in Crowds

Fernando

Denman

Sridharan

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

This paper presents a novel deep learning framework for human trajectory prediction and detecting social group membership in crowds. We introduce a generative adversarial pipeline which preserves the spatio-temporal structure of the pedestrian's neighbourhood, enabling us to extract relevant attributes describing their social identity. We formulate the group detection task as an unsupervised learning problem, obviating the need for supervised learning of group memberships via hand labeled databases, allowing us to directly employ the proposed framework in different surveillance settings. We evaluate the proposed trajectory prediction and group detection frameworks on multiple public benchmarks, and for both tasks the proposed method demonstrates its capability to better anticipate human sociological behaviour compared to the existing state-of-the-art methods. 1

show abstract

Online Bayesian Non-parametrics for Social Group Detection

Cited by 19 publications

References 24 publications

Joint Individual-Group Modeling for Tracking

Joint Individual-Group Modeling for Tracking

Social Group Discovery from Surveillance Videos: A Data-Driven Approach with Attention-Based Cues

GD-GAN: Generative Adversarial Networks for Trajectory Prediction and Group Detection in Crowds

Contact Info

Product

Resources

About