%0 Conference Proceedings
%T Large-scale Graph Indexing using Binary Embeddings of Node Contexts
%A Pau Riba
%A Josep Llados
%A Alicia Fornes
%A Anjan Dutta
%E C.-L.Liu
%E B.Luo
%E W.G.Kropatsch
%E J.Cheng
%B 10th IAPR-TC15 Workshop on Graph-based Representations in Pattern Recognition
%D 2015
%V 9069
%I Springer International Publishing
%@ 0302-9743
%@ 978-3-319-18223-0
%F Pau Riba2015
%O DAG; 600.061; 602.006; 600.077
%O exported from refbase (http://refbase.cvc.uab.es/show.php?record=2618), last updated on Wed, 16 Jan 2019 09:14:48 +0100
%X Graph-based representations are experiencing a growing usage in visual recognition and retrieval due to their representational power in front of classical appearance-based representations in terms of feature vectors. Retrieving a query graph from a large dataset of graphs has the drawback of the high computational complexity required to compare the query and the target graphs. The most important property for a large-scale retrieval is the search time complexity to be sub-linear in the number of database examples. In this paper we propose a fast indexation formalism for graph retrieval. A binary embedding is defined as hashing keys for graph nodes. Given a database of labeled graphs, graph nodes are complemented with vectors of attributes representing their local context. Hence, each attribute counts the length of a walk of order k originated in a vertex with label l. Each attribute vector is converted to a binary code applying a binary-valued hash function. Therefore, graph retrieval is formulated in terms of finding target graphs in the database whose nodes have a small Hamming distance from the query nodes, easily computed with bitwise logical operators. As an application example, we validate the performance of the proposed methods in a handwritten word spotting scenario in images of historical documents.
%K Graph matching
%K Graph indexing
%K Application in document analysis
%K Word spotting
%K Binary embedding
%U http://www.springer.com/us/book/9783319182230
%U http://refbase.cvc.uab.es/files/RLF2015a.pdf
%U http://dx.doi.org/10.1007/978-3-319-18224-7_21
%P 208-217