Motion Segmentation Using Spectral Clustering on Indian Road Scenes
Abstract
We propose a novel motion segmentation formulation over spatio-temporal depth images obtained from stereo sequences that segments multiple motion models in the scene in an unsupervised manner. The motion segmentation is obtained at frame rates that compete with the speed of the stereo depth computation. This is possible due to a decoupling framework that first delineates spatial clusters and subsequently assigns motion labels to each of these cluster with analysis of a novel motion graph model. A principled computation of the weights of the motion graph that signifies the relative shear and stretch between possible clusters lends itself to a high fidelity segmentation of the motion models in the scene.
Cite
Text
Sandhu et al. "Motion Segmentation Using Spectral Clustering on Indian Road Scenes." European Conference on Computer Vision Workshops, 2018. doi:10.1007/978-3-030-11021-5_42Markdown
[Sandhu et al. "Motion Segmentation Using Spectral Clustering on Indian Road Scenes." European Conference on Computer Vision Workshops, 2018.](https://mlanthology.org/eccvw/2018/sandhu2018eccvw-motion/) doi:10.1007/978-3-030-11021-5_42BibTeX
@inproceedings{sandhu2018eccvw-motion,
title = {{Motion Segmentation Using Spectral Clustering on Indian Road Scenes}},
author = {Sandhu, Mahtab and Upadhyay, Sarthak and Krishna, K. Madhava and Medasani, Shanti},
booktitle = {European Conference on Computer Vision Workshops},
year = {2018},
pages = {676-687},
doi = {10.1007/978-3-030-11021-5_42},
url = {https://mlanthology.org/eccvw/2018/sandhu2018eccvw-motion/}
}