M
Marcello Pelillo
Researcher at Ca' Foscari University of Venice
Publications - 278
Citations - 9670
Marcello Pelillo is an academic researcher from Ca' Foscari University of Venice. The author has contributed to research in topics: Cluster analysis & Clique problem. The author has an hindex of 43, co-authored 259 publications receiving 7993 citations. Previous affiliations of Marcello Pelillo include University of Bari & Yale University.
Papers
More filters
Proceedings ArticleDOI
DOTA: A Large-Scale Dataset for Object Detection in Aerial Images
Gui-Song Xia,Xiang Bai,Jian Ding,Zhen Zhu,Serge Belongie,Jiebo Luo,Mihai Datcu,Marcello Pelillo,Liangpei Zhang +8 more
TL;DR: The Dataset for Object Detection in Aerial Images (DOTA) as discussed by the authors is a large-scale dataset of aerial images collected from different sensors and platforms and contains objects exhibiting a wide variety of scales, orientations, and shapes.
Book ChapterDOI
The maximum clique problem
TL;DR: A survey of results concerning algorithms, complexity, and applications of the maximum clique problem is presented and enumerative and exact algorithms, heuristics, and a variety of other proposed methods are discussed.
Journal ArticleDOI
Matching hierarchical structures using association graphs
TL;DR: It is proved that, in the new formulation, there is a one-to-one correspondence between maximal cliques and maximal subtree isomorphisms, which allows the tree matching problem to be cast as an indefinite quadratic program using the Motzkin-Straus theorem.
Journal ArticleDOI
Dominant Sets and Pairwise Clustering
TL;DR: A correspondence between dominant sets and the extrema of a quadratic form over the standard simplex is established, thereby allowing the use of straightforward and easily implementable continuous optimization techniques from evolutionary game theory.
Journal ArticleDOI
An iterative pruning algorithm for feedforward neural networks
TL;DR: A new pruning method is developed, based on the idea of iteratively eliminating units and adjusting the remaining weights in such a way that the network performance does not worsen over the entire training set.