Articles producció científica> Enginyeria Informàtica i Matemàtiques

Interactive online learning for graph matching using active strategies

  • Identification data

    Identifier: imarina:6961617
    Authors:
    Conte, DonatelloSerratosa, Francesc
    Abstract:
    © 2020 Elsevier B.V. In some pattern recognition applications, objects are represented by attributed graphs, in which nodes represent local parts of the objects and edges represent relationships between these local parts. In this framework, the comparison between objects is performed through the distance between attributed graphs. Usually, this distance is a linear equation defined by some cost functions on the nodes and on the edges of both attributed graphs. In this paper, we present an online, active and interactive method for learning these cost functions, which works as follows. Graphs are provided to the learning algorithm by pairs in a sequential order (online). Then, a correspondence between them is computed, and there is a strategy that, given the current pair of graphs and the computed correspondence, proposes which node-to-node mapping would most contribute to the learning process (active). Finally, the human can correct some node-to-node mappings if the human thinks they are wrong (interactive). This is the first learning method applied to graph matching that has the following two features: Being an online method and being active and interactive. These properties make our method useful in the cases that data does not arrive at once and when the human can interact on the system. Thus, given some human interactions the method would have to tend to gradually increase its accuracy. The results show that with few interactions, we achieve better results than the offline learning state of the art methods that are currently available.
  • Others:

    Author, as appears in the article.: Conte, Donatello; Serratosa, Francesc
    Department: Enginyeria Informàtica i Matemàtiques
    URV's Author/s: Serratosa Casanelles, Francesc d'Assís
    Keywords: Optimality Online learning Models Human interaction Graph matching Edit distance Costs functions Costs Cooperative pose estimation Computation Assignment Algorithms Active learning
    Abstract: © 2020 Elsevier B.V. In some pattern recognition applications, objects are represented by attributed graphs, in which nodes represent local parts of the objects and edges represent relationships between these local parts. In this framework, the comparison between objects is performed through the distance between attributed graphs. Usually, this distance is a linear equation defined by some cost functions on the nodes and on the edges of both attributed graphs. In this paper, we present an online, active and interactive method for learning these cost functions, which works as follows. Graphs are provided to the learning algorithm by pairs in a sequential order (online). Then, a correspondence between them is computed, and there is a strategy that, given the current pair of graphs and the computed correspondence, proposes which node-to-node mapping would most contribute to the learning process (active). Finally, the human can correct some node-to-node mappings if the human thinks they are wrong (interactive). This is the first learning method applied to graph matching that has the following two features: Being an online method and being active and interactive. These properties make our method useful in the cases that data does not arrive at once and when the human can interact on the system. Thus, given some human interactions the method would have to tend to gradually increase its accuracy. The results show that with few interactions, we achieve better results than the offline learning state of the art methods that are currently available.
    Thematic Areas: Software Matemática / probabilidade e estatística Management information systems Interdisciplinar Information systems and management Información y documentación Engenharias iv Engenharias iii Economia Computer science, artificial intelligence Ciencias sociales Ciências biológicas i Ciência da computação Astronomia / física Artificial intelligence Administração pública e de empresas, ciências contábeis e turismo
    licence for use: https://creativecommons.org/licenses/by/3.0/es/
    ISSN: 0950-7051
    Author's mail: francesc.serratosa@urv.cat
    Author identifier: 0000-0001-6112-5913
    Record's date: 2024-10-12
    Journal volume: 205
    Papper version: info:eu-repo/semantics/acceptedVersion
    Link to the original source: https://www.sciencedirect.com/science/article/abs/pii/S0950705120304585?via%3Dihub
    Licence document URL: https://repositori.urv.cat/ca/proteccio-de-dades/
    Papper original source: Knowledge-Based Systems. 205 (106275): 106275-
    APA: Conte, Donatello; Serratosa, Francesc (2020). Interactive online learning for graph matching using active strategies. Knowledge-Based Systems, 205(106275), 106275-. DOI: 10.1016/j.knosys.2020.106275
    Article's DOI: 10.1016/j.knosys.2020.106275
    Entity: Universitat Rovira i Virgili
    Journal publication year: 2020
    Publication Type: Journal Publications
  • Keywords:

    Artificial Intelligence,Computer Science, Artificial Intelligence,Information Systems and Management,Management Information Systems,Software
    Optimality
    Online learning
    Models
    Human interaction
    Graph matching
    Edit distance
    Costs functions
    Costs
    Cooperative pose estimation
    Computation
    Assignment
    Algorithms
    Active learning
    Software
    Matemática / probabilidade e estatística
    Management information systems
    Interdisciplinar
    Information systems and management
    Información y documentación
    Engenharias iv
    Engenharias iii
    Economia
    Computer science, artificial intelligence
    Ciencias sociales
    Ciências biológicas i
    Ciência da computação
    Astronomia / física
    Artificial intelligence
    Administração pública e de empresas, ciências contábeis e turismo
  • Documents:

  • Cerca a google

    Search to google scholar