Presentation 2014-10-09
Reduction and Synthesis of Feature Values for High Quality of Experience Image Completion
Mariko ISOGAWA, Dan MIKAMI, Akira KOJIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes a new framework to repair damaged images and videos not in an original feature space but in a projected lower dimensional feature space to achieve effective completion even when a damaged area includes complex structures and changes in color. Photos and videos that we take sometimes include unintended objects, e.g., hiding main subjects or distracting background objects. Many techniques for removing and restoring such missing/damaged areas have been proposed; most of them assume that such areas can be recovered from an image/video database. However, these previous technique's assumptions does not hold when the restoring target contains complex structures or color distribution. To solve this problem, we propose a novel framework that consists of two steps. First, an image/video will be repaired in a projected low dimensional feature space. Then an inverse-projection is used to generate the restored image. With our proposed framework, missing areas with complex structure or color can be well repaired.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Completion / Inpainting / Diminished Reality
Paper # MVE2014-37
Date of Issue

Conference Information
Committee MVE
Conference Date 2014/10/2(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Media Experience and Virtual Environment (MVE)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Reduction and Synthesis of Feature Values for High Quality of Experience Image Completion
Sub Title (in English)
Keyword(1) Completion
Keyword(2) Inpainting
Keyword(3) Diminished Reality
1st Author's Name Mariko ISOGAWA
1st Author's Affiliation NTT Media Intelligence Laboratories()
2nd Author's Name Dan MIKAMI
2nd Author's Affiliation NTT Media Intelligence Laboratories
3rd Author's Name Akira KOJIMA
3rd Author's Affiliation NTT Media Intelligence Laboratories
Date 2014-10-09
Paper # MVE2014-37
Volume (vol) vol.114
Number (no) 239
Page pp.pp.-
#Pages 6
Date of Issue