/Representing Pairwise Spatial and Temporal Relations for Action Recognition

Representing Pairwise Spatial and Temporal Relations for Action Recognition

Pyry K. Matikainen, Martial Hebert and Rahul Sukthankar
Conference Paper, European Conference on Computer Vision 2010 (ECCV 2010), September, 2010

Download Publication (PDF)

Copyright notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author’s copyright. These works may not be reposted without the explicit permission of the copyright holder.


The popular bag-of-words paradigm for action recognition tasks is based on building histograms of quantized features, typically at the cost of discarding all information about relationships between them. However, although the bene cial nature of including these relationships seems obvious, in practice nding good representations for feature relationships in video is di cult. We propose a simple and computationally efficient method for expressing pairwise relationships between quantized features that combines the power of discriminative representations with key aspects of Na ve Bayes. We demonstrate how our technique can augment both appearance- and motion-based features, and that it significantly improves performance on both types of features.

BibTeX Reference
author = {Pyry K. Matikainen and Martial Hebert and Rahul Sukthankar},
title = {Representing Pairwise Spatial and Temporal Relations for Action Recognition},
booktitle = {European Conference on Computer Vision 2010 (ECCV 2010)},
year = {2010},
month = {September},
editor = {K. Daniilidis, P. Maragos, N. Paragios},
keywords = {action recognition, spatial relationships},