FlowNet: Learning optical flow with convolutional networks

Dosovitskiy A, Fischer P, Ilg E, Haeusser P, Hazirbas C, Golkov V, Van Der Smagt P, Cremers D, Brox T (2015)

Publication Type: Conference contribution

Publication year: 2015

Publisher: Institute of Electrical and Electronics Engineers Inc.

Book Volume: 2015 International Conference on Computer Vision, ICCV 2015

Pages Range: 2758-2766

Conference Proceedings Title: Proceedings of the IEEE International Conference on Computer Vision

Event location: Santiago, CHL

ISBN: 9781467383912

DOI: 10.1109/ICCV.2015.316

Abstract

Convolutional neural networks (CNNs) have recently been very successful in a variety of computer vision tasks, especially on those linked to recognition. Optical flow estimation has not been among the tasks CNNs succeeded at. In this paper we construct CNNs which are capable of solving the optical flow estimation problem as a supervised learning task. We propose and compare two architectures: a generic architecture and another one including a layer that correlates feature vectors at different image locations. Since existing ground truth data sets are not sufficiently large to train a CNN, we generate a large synthetic Flying Chairs dataset. We show that networks trained on this unrealistic data still generalize very well to existing datasets such as Sintel and KITTI, achieving competitive accuracy at frame rates of 5 to 10 fps.

Involved external institutions

Albert-Ludwigs-Universität Freiburg DE Germany (DE)

How to cite

APA:

Dosovitskiy, A., Fischer, P., Ilg, E., Haeusser, P., Hazirbas, C., Golkov, V.,... Brox, T. (2015). FlowNet: Learning optical flow with convolutional networks. In Proceedings of the IEEE International Conference on Computer Vision (pp. 2758-2766). Santiago, CHL: Institute of Electrical and Electronics Engineers Inc..

MLA:

Dosovitskiy, Alexey, et al. "FlowNet: Learning optical flow with convolutional networks." Proceedings of the 15th IEEE International Conference on Computer Vision, ICCV 2015, Santiago, CHL Institute of Electrical and Electronics Engineers Inc., 2015. 2758-2766.

BibTeX: Download