ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 Object-centered View Synthesis using Learning-based Image Inpainting
Cited 0 time in scopus Download 1 time Share share facebook twitter linkedin kakaostory
저자
신홍창, 이광순, 음호민, 서정일
발행일
201911
출처
International Conference on 3D Systems and Applications (3DSA) 2019, pp.154-156
협약과제
19HR2700, [전문연구실] 이머시브 미디어 전문연구실, 서정일
초록
This paper presents an object-centered view synthesis technique using multilayer concept. we divide the image into multiple layers based on depth information and then provide different motion parallaxes for each layer depending on the depth. When the disocclusion region appears due to motion parallax, the uncovered region is filled by using a learning-based image inpainting.
KSP 제안 키워드
Depth information, Image inpainting, Learning-based, Motion parallax, Synthesis techniques, multiple layers, view synthesis