We present a novel approach for discovering human interactions in videos. Activity understanding techniques usually require a large number of labeled examples, which are not available in many practical cases. Here, we focus on recovering semantically meaningful clusters of human-human and human-object interaction in an unsupervised fashion. A new iterative solution is introduced based on Maximum Margin Clustering (MMC), which also accepts user feedback to refine clusters. This is achieved by formulating the whole process as a unified constrained latent max-margin clustering problem. Extensive experiments have been carried out over three challenging datasets, Collective Activity, VIRAT, and UT-interaction. Empirical results demonstrate that the proposed algorithm can efficiently discover perfect semantic clusters of human interactions with only a small amount of labeling effort.
@article{arxiv.1502.03851,
title = {Discovering Human Interactions in Videos with Limited Data Labeling},
author = {Mehran Khodabandeh and Arash Vahdat and Guang-Tong Zhou and Hossein Hajimirsadeghi and Mehrsan Javan Roshtkhari and Greg Mori and Stephen Se},
journal= {arXiv preprint arXiv:1502.03851},
year = {2015}
}