TY - GEN
T1 - A Design of Real-Time Style-Transfer Operations in a Game Engine
AU - Park, Haechan
AU - Baek, Nakhoon
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025
Y1 - 2025
N2 - Style transfer operations are the process for transferring a specific style of an image to the target image, while preserving the visual contents of the target image. Currently, the style transfer operations are usually implemented with machine learning techniques. Our goal is to finally achieve real-time style-transfer operations, as the special effect tools on the game engines, which is typically composed of various modules required to develop games. To efficiently implement the style transfer operations, as a special kind of machine-learning-related operations, we need to maintain efficient data management methods and also efficient handling of neural-network related operations. In this work, we designed a separate data processing method with Python modules, which are specialized for inferencing, to preserve the performance of the underlying game engine, as much as possible.
AB - Style transfer operations are the process for transferring a specific style of an image to the target image, while preserving the visual contents of the target image. Currently, the style transfer operations are usually implemented with machine learning techniques. Our goal is to finally achieve real-time style-transfer operations, as the special effect tools on the game engines, which is typically composed of various modules required to develop games. To efficiently implement the style transfer operations, as a special kind of machine-learning-related operations, we need to maintain efficient data management methods and also efficient handling of neural-network related operations. In this work, we designed a separate data processing method with Python modules, which are specialized for inferencing, to preserve the performance of the underlying game engine, as much as possible.
KW - game engine
KW - real-time processing
KW - style transfer
UR - http://www.scopus.com/inward/record.url?scp=85218445884&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-77389-1_33
DO - 10.1007/978-3-031-77389-1_33
M3 - Conference contribution
AN - SCOPUS:85218445884
SN - 9783031773884
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 424
EP - 435
BT - Advances in Visual Computing - 19th International Symposium, ISVC 2024, Proceedings
A2 - Bebis, George
A2 - Patel, Vishal
A2 - Gu, Jinwei
A2 - Panetta, Julian
A2 - Gingold, Yotam
A2 - Johnsen, Kyle
A2 - Arefin, Mohammed Safayet
A2 - Dutta, Soumya
A2 - Biswas, Ayan
PB - Springer Science and Business Media Deutschland GmbH
T2 - 19th International Symposium on Visual Computing, ISVC 2024
Y2 - 21 October 2024 through 23 October 2024
ER -