Construction of a theory deriving super-versatility cross-media bases for breaking limitations in various image processing fields

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K12023
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Multimedia database
• Research Institution: Hokkaido University
• Principal Investigator: HASEYAMA Miki 北海道大学, 情報科学研究科, 教授 (00218463)
• Co-Investigator(Kenkyū-buntansha): 小川 貴弘 北海道大学, 情報科学研究科, 准教授 (20524028)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: マルチメディア情報処理

Outline of Final Research Achievements

This study constructed a new theory for deriving super-versatility cross-media bases to break limitations in various research fields related to image processing such as image coding, restoration, recognition, retrieval and recommendation, etc. Specifically, by collaboratively analyzing both of image data and their corresponding semantic information, sematic analysis of each area in images was realized to simultaneously derive new bases for their optimal approximation. Since these bases included high versatilities, they could be applied to various kinds of research fields related to image processing, and the performance improvement were realized. Therefore, in this study, we actually confirmed that our new approach provided breakthroughs for each research field by using the derived cross-media bases.

Free Research Field

画像処理