Enhanced optical flow estimation via multiscale kernel selection and super-resolution integration
The Visual Computer · 2026-01 · 42 (1)
DOI: 10.1007/s00371-025-04198-9 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Visual Computer, online erschienen am 01.12.2025. Verfasst von Haoxin Guo und 2 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 53 Literaturangaben. Gefördert von: Jilin Provincial Scientific and Technological Development Program, Jilin Provincial Institutions Science and Technology Commissioned.
Abstract (Originalsprache)
Optical flow estimation, pivotal in computer vision, captures motion between adjacent frames. Traditional methods like RAFT suffer from limited feature capture and adaptability in weak texture regions. We propose a refined optical flow estimation network integrating multiscale kernel selection and super-resolution techniques. A parameter-free Swift 3D weighted attention mechanism dynamically adjusts feature importance, while a super-resolution module enhances fine-grained feature capture. A context network with multiscale kernel selectivity expands the receptive field, improving adaptability. Experimental results demonstrate significant accuracy improvements on datasets like Sintel and FlyingChairs, showcasing the method's effectiveness in complex scenes. Code location https://github.com/yifanna/PS3D-flow.git.
Förderung
- Jilin Provincial Scientific and Technological Development Program
- Jilin Provincial Institutions Science and Technology Commissioned
Zitieren
APA
Guo, H., Wang, Y., & Guo, X. (2026). Enhanced optical flow estimation via multiscale kernel selection and super-resolution integration. The Visual Computer, 42(1). https://doi.org/10.1007/s00371-025-04198-9
Vancouver
Guo H, Wang Y, Guo X. Enhanced optical flow estimation via multiscale kernel selection and super-resolution integration. The Visual Computer. 2026;42(1). doi:10.1007/s00371-025-04198-9
BibTeX
@article{Guo2026,
author = {Guo, Haoxin and Wang, Yifan and Guo, Xiaobo},
title = {Enhanced optical flow estimation via multiscale kernel selection and super-resolution integration},
journal = {The Visual Computer},
year = {2026},
volume = {42},
number = {1},
doi = {10.1007/s00371-025-04198-9}
}
EUVIMED