A General Framework for Motion Segmentation: Independent, Articulated, Rigid, Non-rigid, Degenerate and Non-degenerate

By Jingyu Yan, Marc Pollefeys. European Conference on Computer Vision, 2006

583 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/7785c92d471ef3d2ec132f3facf2c1f7ec7e56a5

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Visual SLAM and Structure from Motion in Dynamic Environments (2018, methodology): “While the previous works assumed that the motions are rigid, Yan and Pollefeys [177] proposed a general framework called Local Subspace Affinity (LSA) for independent, articulated, rigid, nonrigid, degenerate, and nondegenerate motions.”
Robust Subspace Clustering (2013, methodology): “Many other methods apply spectral clustering to a specially constructed graph [2, 4, 13, 16, 17, 25, 55, 58].”
Object Segmentation by Long Term Analysis of Point Trajectories (2010, methodology): “When running ALC we randomly subsampled the tracks by a factor 16 to have more tractable computation times. tracks time our method 15486 497s GPCA 12060 2963s LSA 12060 38614s RANSAC 12060 15s ALC 957 22837s Apart from the numbers of the proposed technique we also report numbers for Generalized PCA ”
A Continuous Occlusion Model for Road Scene Understanding (2016, methodology): “For moving objects, this is within the purview of motion segmentation, which has been approached through algebraic factorization methods [4, 27, 26], statistical methods [11, 9, 20] and clustering methods [31, 8].”
Subspace Clustering by Block Diagonal Representation (2018, background): “The way of affinity matrix construction by using the typical Gaussian kernel, or other local information basedmethods, e.g., Local Subspace Affinity (LSA) [40], may not be a good choice for subspace clustering since the data points in a union of subspaces may be distributed arbitrarily but not neces”
Figure-ground segmentation improves handled object recognition in egocentric video (2010, background): “Sparse features are easier to track or match, and feature tracks are often used in motion segmentation (e.g. [9] [24] [28]).”
Motion Segmentation in the Presence of Outlying, Incomplete, or Corrupted Trajectories (2010, methodology): “In these experiments, we compare the robustness to sample outliers of ALC7 and LSA [31], a spectral clustering-based (a) Misclassification rates (in percent) and (b) average computation times.”
Sparse Subspace Clustering: Algorithm, Theory, and Applications (2013, methodology): “LSA and SCC do not have an explicit way to deal with corrupted data.”
Sparse Subspace Clustering: Algorithm, Theory, and Applications (2012, background): “Therefore, there is a need for having clustering algorithms that take into account the multisubspace structure of the data.”
Coarse-to-fine low-rank structure-from-motion (2008, methodology): “Existing approaches are based either on inspecting the eigenvalues of the data matrix [14] or on model selection criteria such as BIC [1] or GRIC [3].”
Learning to recognize objects in egocentric activities (2011, background): “However, problem is much harder for a moving camera and is usually approached by motion segmentation given sparse feature correspondences (e.g. [25]).”
Robust Recovery of Subspace Structures by Low-Rank Representation (2010, methodology): “The procedure of usingE∗ to perform outlier detection is the same as LRR. 4) Some other Methods:We also consider for comparison some previous subspace segmentation methods, including Random Sample Consensus (RANSAC) [10], Generalized PCA (GPCA) [24], Local Subspace Analysis (LSA) [2], Agglomerative ”
Multiframe Motion Segmentation with Missing Data Using PowerFactorization and GPCA (2008, unknown): “For the sequences taken from (Sugaya and Kanatani 2004; Vidal et al. 2006; Yan and Pollefeys 2006), the point trajectories were provided in the respective datasets.”
Ieee Transactions on Pattern Analysis and Machine Intelligence Segmentation of Moving Objects by Long Term Video Analysis (2014, unknown): “More recent methods model the (linear) dependency of the data samples [72], [26], i.e., motion segmentation is cast as the problem of segmenting samples drawn from a union of linear (or affine) subspaces.”
Smooth Representation Clustering (2014, methodology): “A direct way is to compute afﬁnity matrix from existing algebraic methods [ 21] or by deﬁn-ing a point-to-subspace or subspace-to-subspace distance metric [30].”