Learning Neural Velocity Fields from Dynamic 3D Scenes via Edge-Aware Ray Sampling

Sota Ito, Yoshikazu Hayashi, Hiroaki Aizawa, Kunihito Kato

2025

Abstract

Neural Velocity Fields enables future frame extrapolation by learning not only the geometry and appearance but also the velocity of dynamic 3D scenes, by incorporating physics-based constraints. While the divergence theorem employed in NVFi enforces velocity continuity, it also inadvertently imposes continuity at the boundaries between dynamic objects and background regions. Consequently, the velocities of dynamic objects are reduced by the influence of background regions with zero velocity, which diminishes the quality of extrapolated frames. In our proposed method, we identify object boundaries based on geometric information extracted from NVFi and apply the divergence theorem exclusively to non-boundary regions. This approach allows for more accurate learning of velocities, enhancing the quality of both interpolated and extrapolated frames. Our experiments on the Dynamic Object Dataset demonstrated a 1.6% improvement in PSNR [dB] for interpolated frames and a 0.8% improvement for extrapolated frames.

Paper Citation

in Harvard Style

Ito S., Hayashi Y., Aizawa H. and Kato K. (2025). Learning Neural Velocity Fields from Dynamic 3D Scenes via Edge-Aware Ray Sampling. In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP; ISBN 978-989-758-728-3, SciTePress, pages 699-706. DOI: 10.5220/0013233600003912

in Bibtex Style

@conference{visapp25,
author={Sota Ito and Yoshikazu Hayashi and Hiroaki Aizawa and Kunihito Kato},
title={Learning Neural Velocity Fields from Dynamic 3D Scenes via Edge-Aware Ray Sampling},
booktitle={Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP},
year={2025},
pages={699-706},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013233600003912},
isbn={978-989-758-728-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP
TI - Learning Neural Velocity Fields from Dynamic 3D Scenes via Edge-Aware Ray Sampling
SN - 978-989-758-728-3
AU - Ito S.
AU - Hayashi Y.
AU - Aizawa H.
AU - Kato K.
PY - 2025
SP - 699
EP - 706
DO - 10.5220/0013233600003912
PB - SciTePress