Learning flexible sprites in video layers

Jojić, Nebojša; Frey, Brendan J.

doi:10.1109/cvpr.2001.990476

Cited by 182 publications

(177 citation statements)

References 7 publications

Supporting

Mentioning

177

Contrasting

Order By: Relevance

“…where p(x|j, v) is given as in (3) with f and π indexed by v. Notice how the equation (5) relates to equation (3). Clearly now p(x|j) is a mixture model of the type of model given in (3) so that each mixture component is associated with a visual aspect.…”

Section: Incorporating Multiple Viewpointsmentioning

confidence: 99%

“…If there are L possible objects, and there are J transformations that any one object can undergo, then we will need to consider O(J L ) combinations to explain any image. Jojic and Frey [3] tackled this problem by using a variational inference scheme searching over all transformations simultaneously, while Williams and Titsias [4] developed a sequential approach using robust statistics which searches over the transformations of one object at each time. Both these methods do not require a video sequence and can work on unordered sets of images.…”

Section: Introductionmentioning

confidence: 99%

“…Jojic and Frey [3] provided a principled generative probabilistic framework for learning a layered model allowing transparency between the layers. Williams and Titsias [4] developed a similar model where layers strictly combine by occlusion.…”

Section: Introductionmentioning

confidence: 99%

“…[1], [2]. Jojic and Frey [3] provide a generative probabilistic model framework for this task, but their algorithm is slow as it needs to search over discretized transformations (e.g. translations, or affines) for each layer simultaneously.…”

mentioning

confidence: 99%

See 3 more Smart Citations

Sequential Learning of Layered Models from Video

Titsias

Williams

2006

Toward Category-Level Object Recognition

View full text Add to dashboard Cite

Abstract. A popular framework for the interpretation of image sequences is the layers or sprite model, see e.g. [1], [2]. Jojic and Frey [3] provide a generative probabilistic model framework for this task, but their algorithm is slow as it needs to search over discretized transformations (e.g. translations, or affines) for each layer simultaneously. Exact computation with this model scales exponentially with the number of objects, so Jojic and Frey used an approximate variational algorithm to speed up inference. Williams and Titsias [4] proposed an alternative sequential algorithm for the extraction of objects one at a time using a robust statistical method, thus avoiding the combinatorial explosion. In this chapter we elaborate on our sequential algorithm in the following ways: Firstly, we describe a method to speed up the computation of the transformations based on approximate tracking of the multiple objects in the scene. Secondly, for sequences where the motion of an object is large so that different views (or aspects) of the object are visible at different times in the sequence, we learn appearance models of the different aspects. We demonstrate our method on four video sequences, including a sequence where we learn articulated parts of a human body.

show abstract

Section: Incorporating Multiple Viewpointsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Sequential Learning of Layered Models from Video

Titsias

Williams

2006

Toward Category-Level Object Recognition

View full text Add to dashboard Cite

show abstract

“…Later, Irani showed that flow fields of a rigid scene reside in a low-dimensional subspace and constrained the flow field to reduce the noise in the estimate [16]. The flexible sprites approach of Jojic and Frey automatically learns multiple layers using probabilistic 2D appearance maps in an expectation maximization (EM) framework [19], but is limited to stationary camera scenarios. Niyogi, Adelson and Bergen [1,25] also present methods to detect motion boundaries using oriented spatiotemporal energy models that detect surface texture accretion and deletion.…”

Section: Previous Approaches To Video Segmentationmentioning

confidence: 99%

Automatic video segmentation using spatiotemporal T-junctions

Apostoloff¹,

Fitzgibbon²

2006

Procedings of the British Machine Vision Conference 2006

View full text Add to dashboard Cite

The problem of figure-ground segmentation is of great importance in both video editing and visual perception tasks. Classical video segmentation algorithms approach the problem from one of two perspectives. At one extreme, global approaches constrain the camera motion to simplify the image structure. At the other extreme, local approaches estimate motion in small image regions over a small number of frames and tend to produce noisy signals that are difficult to segment. With recent advances in image segmentation showing that sparse information is often sufficient for figureground segmentation it seems surprising then that with the extra temporal information of video, an unconstrained automatic figure-ground segmentation algorithm still eludes the research community. In this paper we present an automatic video segmentation algorithm that is intermediate between these two extremes and uses spatiotemporal features to regularize the segmentation. Detecting spatiotemporal T-junctions that indicate occlusion edges, we learn an occlusion edge model that is used within a colour contrast sensitive MRF to segment individual frames of a video sequence. T-junctions are learnt and classified using a support vector machine and a Gaussian mixture model is fitted to the (foreground, background) pixel pairs sampled from the detected T-junctions. Graph cut is then used to segment each frame of the video showing that sparse occlusion edge information can automatically initialize the video segmentation problem.

show abstract

Interplay Between Visual and Audio Scene Analysis

Xiong

Huang

Speech Separation by Humans and Machines

View full text Add to dashboard Cite

Learning flexible sprites in video layers

Cited by 182 publications

References 7 publications

Sequential Learning of Layered Models from Video

Sequential Learning of Layered Models from Video

Automatic video segmentation using spatiotemporal T-junctions

Interplay Between Visual and Audio Scene Analysis

Contact Info

Product

Resources

About