Structure from Motion

Carefully integrating physical constraints can dramatically enhance learning leading to very data efficient solutions. Specifically, we study how motion facilitates visual information processing for machine learning. Using vision sensors like RGB cameras, drone-mounted cameras, and event cameras, we develop learning mechanisms are capable of acquiring an understanding of the scene's structure from minimal data - e.g. scene depth, object distances, camera and object motions.

N. Cai and P. Bideau

Active Event-Alignment for Monocular Distance Estimation

Proceedings IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Jan. 2025
DOI: 10.1109/WACV61041.2025.00245

paper, code

P. Bideau, E. Learned-Miller, C. Schmid, and K. Alahari

The Right Spin: Learning Object Motion from Rotation-Compensated Flow Fields

International Journal of Computer Vision, Jan. 2024
DOI: 10.1007/s11263-023-01859-x

paper

M. Halawa, O. Hellwich, and P. Bideau

Action-Based Contrastive Learning for Trajectory Prediction

European Conference on Computer Vision (ECCV), Oct. 2022
DOI: 10.1007/978-3-031-19842-7_9

paper

C. Gu, E. Learned-Miller, D. Sheldon, G. Gallego, and P. Bideau

The Spatio Temporal Poisson Point Process: A Simple Model for the Alignment of Event Camera Data

Proceedings IEEE/CVF International Conference on Computer Vision (ICCV), Jan. 2021
DOI: 10.1109/ICCV48922.2021.01324

paper, code

P. Bideau, R.R. Menon, and E. Learned-Miller

MoA-Net: Self-Supervised Motion Segmentation

European Conference on Computer Vision Workshops (ECCVW), Sept. 2019
DOI: 10.1007/978-3-030-11024-6_55

paper, slides

P. Bideau, A. RoyChowdhury, R.R. Menon, and E. Learned-Miller

The Best of Both Worlds: Combining CNNs and Geomtric Constraints for Hierarchical Motion Segmentation

Proceedings IEEE/CVF International Conference on Computer Vision and Pattern Recognition (CVPR), June. 2018
DOI: 10.1109/CVPR.2018.00060

paper, code, evaluation code

P. Bideau and E. Learned-Miller

A Detailed Rubric for Motion Segmentation

arXiv preprint, Oct. 2016
arXiv: 1610.10033 [cs.CV]

paper

P. Bideau and E. Learned-Miller

It's Moving! A Probabilistic Model for Causal Motion Segmentation in Moving Camera Videos

European Conference on Computer Vision (ECCV), Oct. 2016
DOI: 10.1007/978-3-319-46484-8_26

paper, code (.zip file 125MB), dataset (.zip file 300MB)

Page updated

Google Sites

Report abuse