Matching mixtures of curves for human action recognition
Lists
Only you see this
Abstract
Highlights•Gaussian mixture model framework for clustering motion curves.•Matching training and probe motion curves using the longest common subsequence.•Reduce the length of the time series using dimensionality reduction techniques.•Categorize a probe video sequence using nearest neighbor classification scheme.
BibTeX
@article{vrigkas2014matching,
author = {Michalis Vrigkas and Vasileios Karavasilis and Christophoros Nikou and Ioannis A. Kakadiaris},
title = {{Matching mixtures of curves for human action recognition}},
journal = {Comput. Vis. Image Underst.},
year = {2014},
url = {https://openreview.net/forum?id=R2QQsTBh7V},
}Loading related content...