Home // ICIW 2013, The Eighth International Conference on Internet and Web Applications and Services // View article


Cross-Media Retrieval for Music by Analyzing Changes of Mood with Delta Function for Detecting Impressive Behaviours

Authors:
Yoshiyuki Kato
Shuichi Kurabayashi

Keywords: Cross Media Retrieval; Multimedia Database; User Interface; Mood Analysis;

Abstract:
This paper proposes a system that retrieves music by accepting a sequence of images as a query representing a change of sentiments in music. The system offers a query model that utilizes image files as a media for describing users’ emotional demands for continuous changes of mood in music. This query model offers two types of delta functions, corresponding to music and images. Each delta function measures continuous changes in the corresponding sequential media. Applying the delta functions to the media data generates the values representing changes of moods. Each delta value is normalized distance in the corresponding metric space, thus, comparison of delta values extracted from heterogeneous media data makes it possible to calculate the cross-media relevance score. As a prototype implementation, we have developed a Web-based cross-media retrieval engine that provides an integrated user interface (UI) to create music queries by novices who may submit only rough and simple information.

Pages: 236 to 239

Copyright: Copyright (c) IARIA, 2013

Publication date: June 23, 2013

Published in: conference

ISSN: 2308-3972

ISBN: 978-1-61208-280-6

Location: Rome, Italy

Dates: from June 23, 2013 to June 28, 2013