Unsupervised Flow Refinement near Motion Boundaries
Shuzhi Yu (Duke University),* Hannah H Kim (Duke University), Shuai Yuan (Duke University), Carlo Tomasi (Duke University)
The 33rd British Machine Vision Conference
Abstract
Unsupervised optical flow estimators based on deep learning have attracted increasing attention due to the cost and difficulty of annotating for ground truth. Although performance measured by average End-Point Error (EPE) has improved over the years, flow estimates are still poorer along motion boundaries (MBs), where the flow is not smooth, as is typically assumed, and where features computed by neural networks are contaminated by multiple motions. To improve flow in the unsupervised settings, we design a framework that detects MBs by analyzing visual changes along boundary candidates and replaces motions close to detections with motions farther away. Our proposed algorithm detects boundaries more accurately than a baseline method with the same inputs and is shown to improve estimates from different flow predictors without additional training.Video
Citation
@inproceedings{Yu_2022_BMVC,
author = {Shuzhi Yu and Hannah H Kim and Shuai Yuan and Carlo Tomasi},
title = {Unsupervised Flow Refinement near Motion Boundaries},
booktitle = {33rd British Machine Vision Conference 2022, {BMVC} 2022, London, UK, November 21-24, 2022},
publisher = {{BMVA} Press},
year = {2022},
url = {https://bmvc2022.mpi-inf.mpg.de/0351.pdf}
}
Copyright © 2022
The British Machine Vision Conference is organised by