Quickshift++: Provably Good Initializations for Sample-Based Mean Shift
Abstract
We provide initial seedings to the Quick Shift clustering algorithm, which approximate the locally high-density regions of the data. Such seedings act as more stable and expressive cluster-cores than the singleton modes found by Quick Shift. We establish statistical consistency guarantees for this modification. We then show strong clustering performance on real datasets as well as promising applications to image segmentation.
Cite
Text
Jiang et al. "Quickshift++: Provably Good Initializations for Sample-Based Mean Shift." International Conference on Machine Learning, 2018.Markdown
[Jiang et al. "Quickshift++: Provably Good Initializations for Sample-Based Mean Shift." International Conference on Machine Learning, 2018.](https://mlanthology.org/icml/2018/jiang2018icml-quickshift/)BibTeX
@inproceedings{jiang2018icml-quickshift,
title = {{Quickshift++: Provably Good Initializations for Sample-Based Mean Shift}},
author = {Jiang, Heinrich and Jang, Jennifer and Kpotufe, Samory},
booktitle = {International Conference on Machine Learning},
year = {2018},
pages = {2294-2303},
volume = {80},
url = {https://mlanthology.org/icml/2018/jiang2018icml-quickshift/}
}