Factorial Hidden Markov Models
Machine Learning - Special issue on learning with probabilistic representations
The Hierarchical Hidden Markov Model: Analysis and Applications
Machine Learning
Fine-Grained Activity Recognition by Aggregating Abstract Object Usage
ISWC '05 Proceedings of the Ninth IEEE International Symposium on Wearable Computers
Predicting Structured Data (Neural Information Processing)
Predicting Structured Data (Neural Information Processing)
Common sense based joint training of human activity recognizers
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
Decomposition in hidden Markov models for activity recognition
MCAM'07 Proceedings of the 2007 international conference on Multimedia content analysis and mining
Deinterleaving Markov processes via penalized ML
ISIT'09 Proceedings of the 2009 IEEE international conference on Symposium on Information Theory - Volume 3
Probabilistic situation recognition for vehicular traffic scenarios
ICRA'09 Proceedings of the 2009 IEEE international conference on Robotics and Automation
An unsupervised approach to activity recognition and segmentation based on object-use fingerprints
Data & Knowledge Engineering
Trading expressivity for efficiency in statistical relational learning: Ph.D. thesis abstract
ACM SIGKDD Explorations Newsletter
Don't fear optimality: sampling for probabilistic-logic sequence models
ILP'09 Proceedings of the 19th international conference on Inductive logic programming
| Hi-index | 0.01 |
Hidden Markov models assume that observations in time series data stem from some hidden process that can be compactly represented as a Markov chain. We generalize this model by assuming that the observed data stems from multiple hidden processes, whose outputs interleave to form the sequence of observations. Exact inference in this model is NP-hard. However, a tractable and effective inference algorithm is obtained by extending structured approximate inference methods used in factorial hidden Markov models. The proposed model is evaluated in an activity recognition domain, where multiple activities interleave and together generate a stream of sensor observations. It is shown to be more accurate than a standard hidden Markov model in this domain.