Video Camera Trajectory Editing with Generative Rendering from Estimated Geometry
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (11) · 8787-8795
DOI: 10.1609/aaai.v40i11.37832 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Junyoung Seo und 11 weiteren Autor:innen.
Abstract (Originalsprache)
We introduce a novel framework for video camera trajectory editing, enabling the re-synthesis of monocular videos along user-defined camera paths. This task is challenging due to its ill-posed nature and the limited multi-view video data for training. Traditional reconstruction methods struggle with extreme trajectory changes, and existing generative models for dynamic novel view synthesis cannot handle in-the-wild videos. Our approach consists of two steps: estimating temporally consistent geometry, and generative rendering guided by this geometry. By integrating geometric priors, the generative model focuses on synthesizing realistic details where the estimated geometry is uncertain. We eliminate the need for extensive 4D training data through a factorized fine-tuning framework that separately trains spatial and temporal components using multi-view image and video data. Our method outperforms baselines in producing plausible videos from novel camera trajectories, especially in extreme extrapolation scenarios on real-world footage.
Zitieren
APA
Seo, J., Han, J., Jung, J., Jin, S., Lee, J., Narihira, T., Fukuda, K., Shibuya, T., Ahn, D., Hu, S., Kim, S., & Mitsufuji, Y. (2026). Video Camera Trajectory Editing with Generative Rendering from Estimated Geometry. Proceedings of the AAAI Conference on Artificial Intelligence, 40(11), 8787-8795. https://doi.org/10.1609/aaai.v40i11.37832
Vancouver
Seo J, Han J, Jung J, Jin S, Lee J, Narihira T, et al. Video Camera Trajectory Editing with Generative Rendering from Estimated Geometry. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(11):8787-8795. doi:10.1609/aaai.v40i11.37832
BibTeX
@article{Seo2026,
author = {Seo, Junyoung and Han, Jisang and Jung, Jaewoo and Jin, Siyoon and Lee, JoungBin and Narihira, Takuya and Fukuda, Kazumi and Shibuya, Takashi and Ahn, Donghoon and Hu, Shoukang and Kim, Seungryong and Mitsufuji, Yuki},
title = {Video Camera Trajectory Editing with Generative Rendering from Estimated Geometry},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {11},
pages = {8787--8795},
doi = {10.1609/aaai.v40i11.37832}
}
EUVIMED