FRIH: Fine-Grained Region-Aware Image Harmonization
Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (5) · 4478-4486
DOI: 10.1609/aaai.v38i5.28246 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Jinlong Peng und 3 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref).
Abstract (Originalsprache)
Image harmonization aims to generate a more realistic appearance of foreground and background for a composite image. All the existing methods perform the same harmonization process for the whole foreground. However, the implanted foreground always contains different appearance patterns. Existing solutions ignore the difference of each color block and lose some specific details. Therefore, we propose a novel global-local two stages framework for Fine-grained Region-aware Image Harmonization (FRIH). In the first stage, the whole input foreground mask is used to make a global coarse-grained harmonization. In the second stage, we adaptively cluster the input foreground mask into several submasks. Each submask and the coarsely adjusted image are concatenated respectively and fed into a lightweight cascaded module, refining the global harmonization result. Moreover, we further design a fusion prediction module to generate the final result, utilizing the different degrees of harmonization results comprehensively. Without bells and whistles, our FRIH achieves a competitive performance on iHarmony4 dataset with a lightweight model.
Zitieren
APA
Peng, J., Luo, Z., Liu, L., & Zhang, B. (2024). FRIH: Fine-Grained Region-Aware Image Harmonization. Proceedings of the AAAI Conference on Artificial Intelligence, 38(5), 4478-4486. https://doi.org/10.1609/aaai.v38i5.28246
Vancouver
Peng J, Luo Z, Liu L, Zhang B. FRIH: Fine-Grained Region-Aware Image Harmonization. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(5):4478-4486. doi:10.1609/aaai.v38i5.28246
BibTeX
@article{Peng2024,
author = {Peng, Jinlong and Luo, Zekun and Liu, Liang and Zhang, Boshen},
title = {FRIH: Fine-Grained Region-Aware Image Harmonization},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2024},
volume = {38},
number = {5},
pages = {4478--4486},
doi = {10.1609/aaai.v38i5.28246}
}
EUVIMED