A Random Walks View of Spectral Segmentation

By Marina Meila, Jianbo Shi. 2001

775 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Unmasking Clever Hans predictors and assessing what machines really learn (2019, methodology): “The SC [53,147,148] used in step (3) is a clustering algorithm with interesting analytic properties, which constitutes the foundation of the analysis pipeline.”
Community detection in graphs (2009, background): “For instance, Meilă and Shi have proven that the normalized cut for a bipartition equals the total probability that a random walker moves from one of the clusters to the other in either sense (Meila and Shi, 2001).”
A Convergent Functional Architecture of the Insula Emerges Across Imaging Modalities (2012, unknown): “Specifically, we used the Meila-Shi (multicut) algorithm (Meila and Shi, 2001), which performs a generalized eigen-decomposition of the normalized Laplacian of a given similarity matrix (here, an eta(2) matrix), then applies the kmeans clustering algorithm to partition the data into K clusters on th”
Event-based analysis of video (2001, methodology): “We then use the Modified-Ncut approach of [24, 12] to cluster the data.”
Learning with Partially Absorbing Random Walks (2012, background): “…leading to a variety of models including PageRank [14] for web page ranking, hitting and commute times [8] for similarity measure between vertices, harmonic functions [20] for semi-supervised learning, diffusion maps [7] for dimensionality reduction, and normalized cuts [12] for clustering.”
Shape representation and classification using the Poisson equation (2004, methodology): “Random walks are used in various vision applications including perceptual grouping and segmentation [3], [24], [38], [42], [55].”
Learning Segmentation by Random Walks (2000, background): “The proof of Proposition 2 is provided in [5].”
CUDA-Based Parallelization of Power Iteration Clustering for Large Datasets (2017, background): “Although they have proved to be robust, spectral clustering methods have a high order of complexity, especially for calculating the eigenvectors and eigenvalues, which can be as high as O(n3) [4].”