ライフログ内の記録画像を対象とした撮影コンテクストに基づく検索  [in Japanese] Context - based Search for Digital Photographs in Life - log  [in Japanese]

Access this Article

Search this Article

Author(s)

Abstract

ライフログは個人の活動を記録したデータである。ライフログを構成するデータの例として、個人が送受信した電子メール、視聴したTV番組、デジタルカメラで撮影した記録画像等があげられる。ライフログの特徴として、同一の活動が複数のメディアで記録されていることがある。それぞれのメディアには特徴があり、表現可能な内容には限界がある。メディアを統合化することにより、同一の活動を様々な視点から捉えることが可能となる。統合化されたライフログは、個別のデータ単体では表現できないコンテクストを相互に表現していると考えられる。本研究では、異種メディアで記録されたライフログを統合化し、コンテクストに基づいた検索を実現することを目的とする。統合化の際に問題となるのは、異種メディアの対応付けを厳密に行うことは困難であり、不確実性を伴う点である。本論文では、この問題に対応するために、推論ネットワークを拡張した検索モデルを提案する。なお、検索対象の例としてデジタルカメラや携帯電話などで撮影された記録画像を想定し、記録画像のコンテクストの推定には、利用者が送受信した電子メールのメッセージを利用する。The life-log is the data that records activities of a person. Examples of life-logs are the e-mail massages that he/she sent and recieved, TV programs that he/she watched, and photographs that he/she took, and so on. One of the most important characteristic of life-log is that the same activities are represented in different media types. Each type has its own characteristic features, and the limitation on its content representation ability. The integration of life-log in different media types enable to represent the same activities from various viewpoints. Integrated life-logs represent contexts each other that are not able to represent in single. The goal of our research is to integrate life-logs in different media and to search them based on their contexts. A problem of the integration is that it is difficult to make correspondences between life-logs in different media types strictly and the correspondences consists uncertainty. In this paper, we introduce a framework based on Bayesian Network. We suppose to search the photographs taken with digital camera based on their recorded contexts, and use e-mail messages for context estimation.

The life-log is the data that records activities of a person. Examples of life-logs are the e-mail massages that he/she sent and recieved, TV programs that he/she watched, and photographs that he/she took, and so on. One of the most important characteristic of life-log is that the same activities are represented in different media types. Each type has its own characteristic features, and the limitation on its content representation ability. The integration of life-log in different media types enable to represent the same activities from various viewpoints. Integrated life-logs represent contexts each other that are not able to represent in single. The goal of our research is to integrate life-logs in different media and to search them based on their contexts. A problem of the integration is that it is difficult to make correspondences between life-logs in different media types strictly and the correspondences consists uncertainty. In this paper, we introduce a framework based on Bayesian Network. We suppose to search the photographs taken with digital camera based on their recorded contexts, and use e-mail messages for context estimation.

Journal

  • IPSJ SIG Notes  

    IPSJ SIG Notes 2004(72(2004-DBS-134)), 285-292, 2004-07-14 

    Information Processing Society of Japan (IPSJ)

References:  17

You must have a user ID to see the references.If you already have a user ID, please click "Login" to access the info.New users can click "Sign Up" to register for an user ID.

Codes

  • NII Article ID (NAID)
    110002911336
  • NII NACSIS-CAT ID (NCID)
    AN10112482
  • Text Lang
    JPN
  • Article Type
    Technical Report
  • ISSN
    09196072
  • NDL Article ID
    7064066
  • NDL Source Classification
    ZM13(科学技術--科学技術一般--データ処理・計算機)
  • NDL Call No.
    Z14-1121
  • Data Source
    CJP  NDL  IPSJ 
Page Top