Segmenting Motions of Different Types by Unsupervised Manifold Clustering

Goh, Alvina; Vidal, René

doi:10.1109/cvpr.2007.383235

Cited by 164 publications

(110 citation statements)

References 20 publications

Supporting

Mentioning

108

Contrasting

Unclassified

Order By: Relevance

“…In general, the motion trajectories from a perspective camera will lie on a nonlinear manifold instead of a linear subspace [8]. However, it is possible to approximate the manifold locally (over a short period of time) with a linear subspace.…”

Section: Subspace Theorymentioning

confidence: 99%

Visual Tracking via Subspace Motion Model

Wang¹,

Zhong²,

Wang³

et al. 2013

Procedings of the British Machine Vision Conference 2013

View full text Add to dashboard Cite

In this paper we propose a novel motion model for visual tracking, which can be used to parameterize instantaneous image motion caused by both object and camera movements. Our approach is inspired by the subspace theory of image motion, that is, for a rigid object imaged by a projective camera, the displacements matrix of its trajectories over a short period of time should approximately lie in a low-dimensional subspace with a certain rank upper bound. We adopt this subspace as the state transition space in particle filtering, by which we can obtain a state vector with alterable number of dimensions. The dimension number as well as the sampling weight for each dimension at each moment can be determined by the rank of the subspace. In this way the particle distribution will be more coincide with the probability of the object state. Based on the subspace motion model, we derive a new visual tracking approach that can handle more complicated cases. The subspace is also used to discriminate new feature points of the object for adaptively updating the motion model. Experimental results and comparisons demonstrate the effectiveness of the proposed method.

show abstract

Section: Subspace Theorymentioning

confidence: 99%

Visual Tracking via Subspace Motion Model

Wang¹,

Zhong²,

Wang³

et al. 2013

Procedings of the British Machine Vision Conference 2013

View full text Add to dashboard Cite

show abstract

“…The method has been applied to partition affine camera motions with outlying and corrupted trajectories (Rao et al 2008). There has also been work to apply manifold learning techniques such as ISOMAP and local linear embedding to cluster data drawn from multiple manifolds (Souvenir and Pless 2007;Goh and Vidal 2007). In the literature of computer vision, many effective methods have been developed based on the above clustering techniques or more views (Ozden et al 2007;Vidal and Hartley 2008).…”

Section: Relations To Previous Workmentioning

confidence: 99%

Robust Algebraic Segmentation of Mixed Rigid-Body and Planar Motions from Two Views

et al. 2010

View full text Add to dashboard Cite

This paper studies segmentation of multiple rigidbody motions in a 3-D dynamic scene under perspective camera projection. We consider dynamic scenes that contain both 3-D rigid-body structures and 2-D planar structures. Based on the well-known epipolar and homography constraints between two views, we propose a hybrid perspective constraint (HPC) to unify the representation of rigidbody and planar motions. Given a mixture of K hybrid perspective constraints, we propose an algebraic process to partition image correspondences to the individual 3-D motions, called Robust Algebraic Segmentation (RAS). Particularly, we prove that the joint distribution of image correspondences is uniquely determined by a set of (2K)-th degree polynomials, a global signature for the union of K motions of possibly mixed type. The first and second deriva- tives of these polynomials provide a means to recover the association of the individual image samples to their respective motions. Finally, using robust statistics, we show that the polynomials can be robustly estimated in the presence of moderate image noise and outliers. We conduct extensive simulations and real experiments to validate the performance of the new algorithm. The results demonstrate that RAS achieves notably higher accuracy than most existing robust motion-segmentation methods, including random sample consensus (RANSAC) and its variations. The implementation of the algorithm is also two to three times faster than the existing methods. The implementation of the algorithm and the benchmark scripts are available at http:// perception.csl.illinois.edu/ras/.

show abstract

“…This motion is either assumed to occur [6], [7], [8] or can be induced by the robot [9], [10]. Although relative motion is a strong cue for segmentation, generating this motion in an unknown pile is oftentimes dangerous and undesirable.…”

Section: A Object Segmentationmentioning

confidence: 99%

Perceiving, learning, and exploiting object affordances for autonomous pile manipulation

et al. 2014

View full text Add to dashboard Cite

Abstract-Autonomous manipulation in unstructured environments presents roboticists with three fundamental challenges: object segmentation, action selection, and motion generation. These challenges become more pronounced when unknown manmade or natural objects are cluttered together in a pile. We present an end-to-end approach to the problem of manipulating unknown objects in a pile, with the objective of removing all objects from the pile and placing them into a bin. Our robot perceives the environment with an RGB-D sensor, segments the pile into objects using non-parametric surface models, computes the affordances of each object, and selects the best affordance and its associated action to execute. Then, our robot instantiates the proper compliant motion primitive to safely execute the desired action. For efficient and reliable action selection, we developed a framework for supervised learning of manipulation expertise. We conducted dozens of trials and report on several hours of experiments involving more than 1500 interactions. The results show that our learning-based approach for pile manipulation outperforms a common sense heuristic as well as a random strategy, and is on par with human action selection.

show abstract

Segmenting Motions of Different Types by Unsupervised Manifold Clustering

Cited by 164 publications

References 20 publications

Visual Tracking via Subspace Motion Model

Visual Tracking via Subspace Motion Model

Robust Algebraic Segmentation of Mixed Rigid-Body and Planar Motions from Two Views

Perceiving, learning, and exploiting object affordances for autonomous pile manipulation

Contact Info

Product

Resources

About