Similar-Video Retrieval via Learned Exemplars and Time-Warped Alignment

Teruki Horie*, Masafumi Moriwaki, Ryota Yokote, Shota Ninomiya, Akihiro Shikano, Yasuo Matsuyama

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    New learning algorithms and systems for retrieving similar videos are presented. Each query is a video itself. For each video, a set of exemplars is machine-learned by new algorithms. Two methods were tried. The first and main one is the time-bound affinity propagation. The second is the harmonic competition which approximates the first. In the similar-video retrieval, the number of exemplar frames is variable according to the length and contents of videos. Therefore, each exemplar possesses responsible frames. By considering this property, we give a novel similarity measure which contains the Levenshtein distance (L-distance) as its special case. This new measure, the M-distance, is applicable to both of global and local alignments for exemplars. Experimental results in view of precision-recall curves show creditable scores in the region of interest.

    Original languageEnglish
    Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    PublisherSpringer Verlag
    Pages85-94
    Number of pages10
    Volume8836
    ISBN (Print)9783319126425
    Publication statusPublished - 2014
    Event21st International Conference on Neural Information Processing, ICONIP 2014 - Kuching
    Duration: 2014 Nov 32014 Nov 6

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume8836
    ISSN (Print)03029743
    ISSN (Electronic)16113349

    Other

    Other21st International Conference on Neural Information Processing, ICONIP 2014
    CityKuching
    Period14/11/314/11/6

    Keywords

    • Exemplar
    • M-distance
    • Numerical label
    • Similar-video retrieval
    • Time-bound affinity propagation

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    Fingerprint

    Dive into the research topics of 'Similar-Video Retrieval via Learned Exemplars and Time-Warped Alignment'. Together they form a unique fingerprint.

    Cite this