Multiclass spectral clustering

By Stella X. Yu, Jianbo Shi. 2003

1,132 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computer Vision and Pattern Recognition, Media Technology, Representation & Self-Supervised Learning

How researchers see this paper

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

In Computational Mechanics

In Computer Vision and Pattern Recognition

In Media Technology

In Representation & Self-Supervised Learning

Citations in the corpus

Local Aggregation for Unsupervised Learning of Visual Embeddings (2019, methodology): “e Table5). Given the need for cluster-based identiﬁcation, a variety of alternative approaches to k-means are theoretically possible, including DBSCAN [16], Afﬁnity Propagation [18], spectral methods [59,62], and gaussian mixtures [56,57]. However, our present context is strongly constrained by the ”
Segmentation using superpixels: A bipartite graph partitioning approach (2012, methodology): “Clustering is performed by applying k-means [22] or certain discretization technique [30].”
Adaptive Affinity Field for Semantic Segmentation (2018, background): “Affinity has a long history in computer vision in defining the pixel relationships [31,32].”
Optimal Clustering Framework for Hyperspectral Band Selection (2018, background): “The NC and normalized association (NA) [42] are defined as”
Fast approximate spectral clustering (2009, methodology): “Different spectral clustering algorithms formalize this partitioning problem in different ways [31, 26, 28, 37].”
Learning Spectral Clustering, With Application To Speech Separation (2006, methodology): “Another natural possibility followed by Yu and Shi (2003) is to compare directly U (or a normalized version thereof) with the indicator matrix E, up to an orthogonal matrix R, which then has to be estimated.”
Learning binary codes for collaborative filtering (2012, background): “Similar ideas have been investigated for spectral clustering [22] and we extend the idea to the context of learning binary codes.”
Learning to segment moving objects in videos (2014, methodology): “Speciﬁ-cally, we discretize the spectral embedding of trajectory motion afﬁnities [40] for varying the number of eigenvectors.”
Hashing with binary autoencoders (2015, methodology): “The latter is based on the optimal discretization algorithm of Yu and Shi (2003), which ﬁnds a rotation of the continuous eigenvectors of a graph Laplacian that makes them as close as possible to a discrete solution, as a postprocessing for spectral clustering.”
Unsupervised Feature Selection Using Nonnegative Spectral Analysis (2012, methodology): “The lo-cal geometrical structure can be utilized by minimizing the following (Shi and Malik 2000; Yu and Shi 2003): where A is a diagonal matrix with A ii = =1 S ij and L = A − 1 / 2 ( A − S ) A − 1 / 2 is the normalized graph Laplacian matrix.”
Superpixel segmentation using Linear Spectral Clustering (2015, methodology): “We compare LSC to eight state of the art superpixel segmentation algorithms including SLIC [1], SEEDS [2], Ncuts [23], Lattice [15], ERS [12], Turbopixel [11], EneOpt1 and EneOpt0 [20].”
A whole brain fMRI atlas generated via spatially constrained spectral clustering (2011, methodology): “NCUT clustering was performed using a Python implementation of the algorithm presented in [Yu and Shi, 2003].”
Image Clustering Using Local Discriminant Models and Global Integration (2010, methodology): “Following [39], instead of directly solving the cluster assignment matrix , we first define the scaled cluster assignment matrix as follows:”
Groupwise whole-brain parcellation from resting-state fMRI data for network node identification (2013, positive): “Yu and Shi (2003) proposed an iterative algorithm called the multiclass spectral clustering to solve the K-way normalized cut.”
Data Fusion and Multicue Data Matching by Diffusion Maps (2006, methodology): “To exemplify their applicability to high-dimensional data processing and learning, we use them to derive a novel high-dimensional data alignment algorithm in Section 2.4.”
A Comprehensive Survey of Clustering Algorithms (2015, background): “For more detailed information about this kind of clustering algorithms, you can refer to [51,84,112–114].”