Hidden Markov models hmms and partially observable Markov decision processes pomdps provide useful tools for modeling dynamical systems. They are particularly useful for represent...
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...
We propose an extractive summarization system with a novel non-generative probabilistic framework for speech summarization. One of the most underutilized features in extractive su...
Pascale Fung, Ricky Ho Yin Chan, Justin Jian Zhang
The BYBLOS continuous speech recognition system is applied to on-line cursive handwriting recognition. By exploiting similarities between on-line cursive handwriting and continuou...
John Makhoul, Thad Starner, Richard M. Schwartz, G...
In this paper, we introduce a Hybrid Hidden Markov Model (HMM) face recognition system. The proposed system contains a low-complexity 2-D HMM-based face recognition (LC 2D-HMM FR)...