Towards complex activity recognition using a Bayesian network-based probabilistic generative framework. (August 2017)

• Record Type: Journal Article
• Authors: Liu, Li; Wang, Shu; Su, Guoxin; Huang, Zi-Gang; Liu, Ming
• Is Part Of: Pattern recognition. Volume 68(2017:Aug.)
• Page Start: 295
• Page End: 309
• Publication Date: 2017-08
• Subjects: Activity recognition -- Bayesian network -- Complex activity -- Probabilistic generative model -- Temporal relation -- Uncertainty; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2017.02.028
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A Bayesian network-based probabilistic generative framework is presented to address diversity and uncertainty in complex activity recognition. The framework introduces the Chinese restaurant process to explicitly characterize the unique configurations of a complex activity. An enhanced model is presented to characterize more temporal relational variabilities than the previous models over our framework. Our models significantly outperform the state-of-the-arts on three benchmark datasets with different challenges. Our approach is robust against the incomplete or incorrect observations of primitive events. Abstract: Complex activity recognition is challenging since a complex activity can be performed in different ways, with each having its own configuration of primitive events and their temporal dependencies. To address such temporal relational variabilities in complex activity recognition, we propose a Bayesian network-based probabilistic generative framework that employs Allen's interval relation network to represent local temporal dependencies in a generative way. By employing the Chinese restaurant process and introducing relation generation constraints, our framework can characterize these unique internal configurations of a particular complex activity as a joint distribution. Three concrete models are implemented based on our framework. Specifically, in this paper we improve two of our previous models and provide an enhanced model to handle temporal relational variabilities in complex activities more efficiently. Empirical evaluations on three benchmark datasets demonstrate the competitiveness of our framework. In particular, it is shown that our models are rather robust against errors caused by the low-level predictions from raw signals.