In-network adaptive cluster enumeration for distributed classification and labeling

Teklehaymanot, Freweyni K.; Muma, Michael; Liu, Jun; Zoubir, Abdelhak M.

doi:10.1109/eusipco.2016.7760288

Cited by 8 publications

(7 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2) Multi-Object Multi-Camera Network Application: The multi-object multi-camera network application [44], [45] depicted in Fig. 5 contains seven cameras that actively monitor a common scene of interest from different view points.…”

Section: Real Data Resultsmentioning

confidence: 99%

“…In this paper, our focus lies on the derivation of a Bayesian model selection criterion for cluster analysis. The estimation of the number of clusters, also called cluster enumeration, has been intensively researched for decades [27]- [45] and a popular approach is to apply the Bayesian Information Criterion (BIC) [29], [31]- [33], [37]- [41], [44]. The BIC finds the large sample limit of the Bayes' estimator which leads to the selection of a model that is a posteriori most probable.…”

Section: Introductionmentioning

confidence: 99%

“…The works in [15], [46] have shown that model selection rules that penalize for model complexity have to be examined carefully before they are applied to specific model selection problems. Nevertheless, despite the widespread use of the BIC for cluster enumeration [29], [31]- [33], [37]- [41], [44], very little effort has been made to check the appropriateness of the original BIC formulation [16] for cluster analysis. One noticeable work towards this direction was made in [38] by providing a more accurate approximation to the marginal likelihood for small sample sizes.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Bayesian Cluster Enumeration Criterion for Unsupervised Learning

Teklehaymanot

Muma

Zoubir

2018

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

We derive a new Bayesian Information Criterion (BIC) by formulating the problem of estimating the number of clusters in an observed data set as maximization of the posterior probability of the candidate models. Given that some mild assumptions are satisfied, we provide a general BIC expression for a broad class of data distributions. This serves as a starting point when deriving the BIC for specific distributions. Along this line, we provide a closed-form BIC expression for multivariate Gaussian distributed variables. We show that incorporating the data structure of the clustering problem into the derivation of the BIC results in an expression whose penalty term is different from that of the original BIC. We propose a two-step cluster enumeration algorithm. First, a model-based unsupervised learning algorithm partitions the data according to a given set of candidate models. Subsequently, the number of clusters is determined as the one associated with the model for which the proposed BIC is maximal. The performance of the proposed two-step algorithm is tested using synthetic and real data sets.

show abstract

Section: Real Data Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Bayesian Cluster Enumeration Criterion for Unsupervised Learning

Teklehaymanot

Muma

Zoubir

2018

IEEE Trans. Signal Process.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The total number of objects (clusters) K n at time instant n is assumed to be known or estimated a priori, e.g. using [15]. Due to the different view points of the cameras, even at the same time instant, the number of objects observed by different cameras differs.…”

Section: Problem Formulationmentioning

confidence: 99%

Adaptive diffusion-based track assisted multi-object labeling in distributed camera networks

Teklehaymanot

Muma

Zoubir

2017

2017 25th European Signal Processing Conference (EUSIPCO)

Self Cite

View full text Add to dashboard Cite

Abstract-The tracking and labeling of multiple objects in multiple cameras is a fundamental task in applications such as video surveillance, autonomous driving, and sports analysis. In an ad-hoc multi-camera network without a fusion center nodes can benefit from local cooperation to solve signal processing tasks, such as distributed image enhancement. A crucial first step for the successful cooperation of neighboring nodes is to answer the question: Who observes what?. In this paper, an adaptive algorithm is proposed that enables cameras with different view points to assign the same identity to the same object across time frames without assuming the availability of camera calibration information or requiring the registration of camera views. Information which is extracted directly from the videos and is shared in the network via a diffusion algorithm is exploited to jointly solve multi-object tracking and labeling problems in a multi-camera network. A real-data use case of pedestrian labeling is provided, which demonstrates that a high labeling accuracy can be achieved in a multi-object multi-camera setup with low video resolution and frequent object occlusions.

show abstract

“…To the best of our knowledge, distributed cluster enumeration has only been addressed in [29], which serves here as a benchmark algorithm. For the single-node case, the question of inferring the number of clusters from the observations has been intensively studied [30]- [53].…”

Section: Introductionmentioning

confidence: 99%

Gravitational Clustering: A simple, robust and adaptive approach for distributed networks

2018

Self Cite

View full text Add to dashboard Cite

Distributed signal processing for wireless sensor networks enables that different devices cooperate to solve different signal processing tasks. A crucial first step is to answer the question: who observes what? Recently, several distributed algorithms have been proposed, which frame the signal/object labelling problem in terms of cluster analysis after extracting sourcespecific features, however, the number of clusters is assumed to be known. We propose a new method called Gravitational Clustering (GC) to adaptively estimate the time-varying number of clusters based on a set of feature vectors. The key idea is to exploit the physical principle of gravitational force between mass units: streaming-in feature vectors are considered as mass units of fixed position in the feature space, around which mobile mass units are injected at each time instant. The cluster enumeration exploits the fact that the highest attraction on the mobile mass units is exerted by regions with a high density of feature vectors, i.e., gravitational clusters. By sharing estimates among neighboring nodes via a diffusion-adaptation scheme, cooperative and distributed cluster enumeration is achieved. Numerical experiments concerning robustness against outliers, convergence and computational complexity are conducted. The application in a distributed cooperative multi-view camera network illustrates the applicability to real-world problems.Index Terms-adaptive distributed clustering, cluster enumeration, robust, outlier, multi device multi task (MDMT), wireless sensor networks, labelling.

show abstract

In-network adaptive cluster enumeration for distributed classification and labeling

Cited by 8 publications

References 11 publications

Bayesian Cluster Enumeration Criterion for Unsupervised Learning

Bayesian Cluster Enumeration Criterion for Unsupervised Learning

Adaptive diffusion-based track assisted multi-object labeling in distributed camera networks

Gravitational Clustering: A simple, robust and adaptive approach for distributed networks

Contact Info

Product

Resources

About