ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper ArtTension: Artistically Extended Generation based on Single-image Learning
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
Jung-Jae Yu, Juwon Lee, Junmo Kim
Issue Date
2021-12
Citation
Conference on Neural Information Processing Systems (NeurIPS) 2021 : Workshop, pp.1-6
Language
English
Type
Conference Paper
Abstract
To enhance efficiency in creative activities that utilize existing art paintings, we propose an artistic outpainting method based on single-image learning. The proposed method expands the image background by learning only the input image, without pre-learning large-capacity images. When recreating existing artworks by transforming them, the simple task of expanding the background is automatically solved, allowing the artist to focus on more creative work. The proposed method can be used in AI collaboration or immersive media art exhibitions using existing paintings.
KSP Keywords
Immersive media art, Large capacity, Single image, creative work