Unsupervised Fine-tuning of Optical Flow for Better Motion Boundary Estimation

Taha Alhersh, Heiner Stuckenschmidt

2019

Abstract

Recently, convolutional neural network (CNN) based approaches have proven to be successful in optical flow estimation in the supervised as well as in the unsupervised training paradigms. Supervised training requires large amounts of training data with task specific motion statistics. Usually, synthetic datasets are used for this purpose. Fully unsupervised approaches are usually harder to train and show weaker performance, although they have access to the true data statistics during training. In this paper we exploit a well-performing pre-trained model and fine-tune it in an unsupervised way using classical optical flow training objectives to learn the dataset specific statistics. Thus, per dataset training time can be reduced from days to less than 1 minute. Specifically, motion boundaries estimated by gradients in the optical flow field can be greatly improved using the proposed unsupervised fine-tuning.

Download


Paper Citation


in Harvard Style

Alhersh T. and Stuckenschmidt H. (2019). Unsupervised Fine-tuning of Optical Flow for Better Motion Boundary Estimation. In Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2019) - Volume 5: VISAPP; ISBN 978-989-758-354-4, SciTePress, pages 776-783. DOI: 10.5220/0007343707760783


in Bibtex Style

@conference{visapp19,
author={Taha Alhersh and Heiner Stuckenschmidt},
title={Unsupervised Fine-tuning of Optical Flow for Better Motion Boundary Estimation},
booktitle={Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2019) - Volume 5: VISAPP},
year={2019},
pages={776-783},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0007343707760783},
isbn={978-989-758-354-4},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2019) - Volume 5: VISAPP
TI - Unsupervised Fine-tuning of Optical Flow for Better Motion Boundary Estimation
SN - 978-989-758-354-4
AU - Alhersh T.
AU - Stuckenschmidt H.
PY - 2019
SP - 776
EP - 783
DO - 10.5220/0007343707760783
PB - SciTePress