In this paper, we analyse the properties of the standard circular hidden Markov model (HMM) on 2D shape recognition. We point out the limitations of the circular HMMs and further ...
For decades, Hidden Markov Models (HMMs) have been the state-of-the-art technique for acoustic modeling despite their unrealistic independence assumptions and the very limited rep...
We describe a new tagging model where the states of a hidden Markov model (HMM) estimated by unsupervised learning are incorporated as the features in a maximum entropy model. Our...
Hidden Markov Models (HMMs) model sequential data in many fields such as text/speech processing and biosignal analysis. Active learning algorithms learn faster and/or better by cl...
The creation of a cognitive perception systems capable of inferring higher-level semantic information from low-level feature and event information for a given type of multimedia co...
Ilias Kolonias, William J. Christmas, Josef Kittle...