J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, pp. 307--323, 2000. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Learning Image Statistics for Bayesian Tracking - Hedvig Sidenbladh Michael (2001) (23 citations) (Correct)
....and previous approaches have used highly simplified noise models. In contrast, the learned models here account for the variation observed in training data. These edge, ridge, and motion models are then combined in a Bayesian framework. Similar in spirit is the tracking work of Sullivan et al. [22, 23] who model the distributions of filter responses for a general background and a particular foreground where the foreground is represented by a generalized template. Given these, they determine if an image patch is background, foreground, or on the boundary by matching the distribution of filter ....
....as ### # #### # # (1) This is the normalized ratio of the probability that the foreground pixels are explained by the person model versus that they are explained by a generic background model. The spatial and temporal statistics of neighboring pixels are unlikely to be independent [23]. We therefore approximate the set ## # # with a randomly sampled subset to approximate pixel independence. The number of samples in the foreground is always the same and covers the visible parts of the human model. The filter responses, are computed from a set of filters that are chosen ....
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, pp. 307--323, 2000.
Learning Image Statistics for Bayesian Tracking - Sidenbladh, Black (2001) (23 citations) (Correct)
....and previous approaches have used highly simplified noise models. In contrast, the learned models here account for the variation observed in training data. These edge, ridge, and motion models are then combined in a Bayesian framework. Similar in spirit is the tracking work of Sullivan et al. [22, 23] who model the distributions of filter responses for a general background and a particular foreground where the foreground is represented by a generalized template. Given these, they determine if an image patch is background, foreground, or on the boundary by matching the distribution of filter ....
....as p(f j ) 1 : 1) This is the normalized ratio of the probability that the foreground pixels are explained by the person model versus that they are explained by a generic background model. The spatial and temporal statistics of neighboring pixels are unlikely to be independent [23]. We therefore approximate the set fx f g with a randomly sampled subset to approximate pixel independence. The number of samples in the foreground is always the same and covers the visible parts of the human model. The filter responses, f , are computed from a set of filters that are chosen to ....
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, pp. 307--323, 2000.
Learning the Statistics of People in Images and Video - Sidenbladh, Black (2001) (7 citations) (Correct)
....is often assumed (Gaussian or some more robust distribution) and is used to derive the likelihood of observing variations from brightness constancy given a predicted motion of the body. These probabilistic formulations have recently been incorporated into Bayesian frameworks for tracking people [4, 8, 17, 38, 43, 44, 45]. These Bayesian methods allow the combination of various image cues, represent ambiguities and multiple hypotheses, and provide a framework for combining new measurements with the previous history of the human motion in a probabilistically sound fashion. The Bayesian methods require a temporal ....
....The use of fixed templates also involves a brightness constancy assumption. However, instead of comparing corresponding image locations between two consecutive frames t and t 1, the image at time t is compared to a reference image at time 0. Templates have been used successfully for face tracking [45], and have also proven suitable for tracking of articulated structures in constrained cases [4, 30] One problem with templates for 3D structures is that the templates are view based. Hence, if the object rotates, the tracking may fail since the system only knows what the object looks like from ....
[Article contains additional citation context not shown here]
Sullivan, J., A. Blake, and J. Rittscher: 2000, 'Statistical foreground modelling for object localisation'. In: D. Vernon (ed.): European Conference on Computer Vision, ECCV, Vol. 2. pp. 307-323.
Automatic Detection and Tracking of Human Motion with a.. - Fablet, Black (2002) (1 citation) (Correct)
....to detect and track in the scene. For these statistical schemes, the key point is to provide appropriate statistical characterization of the entities of interest (foreground) and of the background. Recent work on Bayesian tracking has focused on this problem of foreground background modeling [17,24,26,28]. In this paper, we also consider such a Bayesian approach. Unlike previous work, our main focus is on the definition of appropriate probabilistic models of dynamic information for human motion. As previously mentioned, whereas motion cues provide generic and rich information independent of ....
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, II pp. 307--323, 2000.
Learning Image Statistics for Bayesian Tracking - Hedvig Sidenbladh Michael (2001) (23 citations) (Correct)
No context found.
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, pp. 307--323, 2000.
Learning Image Statistics for Bayesian Tracking - Hedvig Sidenbladh Michael (2001) (23 citations) (Correct)
No context found.
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, pp. 307--323, 2000.
Gibbs Likelihoods for Bayesian Tracking - Stefan Roth Leonid (2004) (2 citations) (Correct)
No context found.
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, 2:307--323, 2000.
Gibbs Likelihoods for Bayesian Tracking - Stefan Roth Leonid (2004) (2 citations) (Correct)
No context found.
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, 2:307--323, 2000.
Gibbs Likelihoods for Bayesian Tracking - Roth, Sigal, Black (2004) (2 citations) (Correct)
No context found.
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, 2:307--323, 2000.
Automatic Detection and Tracking of Human Motion With a.. - Ronan Fablet And (2002) (1 citation) (Correct)
No context found.
J. Sullivan, A. Blake, and J. Rittscher. Statistical foreground modelling for object localisation. ECCV, II pp. 307--323, 2000.
Switching Observation Models for Contour Tracking in Clutter - Ying Wu Gang (2003) (1 citation) (Correct)
No context found.
Josephine Sullivan, Andrew Blake, and Jens Rittscher. Statistical foreground modelling for object localisation. In Proc. of European Conf. on Computer Vision, pages 307--323, 2000.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC