49 search results - page 5 / 10 » Finite State Transducers Approximating Hidden Markov Models |
UAI
2004
14 years 11 months ago
2004
We present a new method for nonlinear prediction of discrete random sequences under minimal structural assumptions. We give a mathematical construction for optimal predictors of s...
105
Voted
ICML
2005
IEEE
15 years 11 months ago
2005
IEEE
This paper explores the issue of recognizing, generalizing and reproducing arbitrary gestures. We aim at extracting a representation that encapsulates only the key aspects of the ...
96
Voted
ECML
2005
Springer
15 years 3 months ago
2005
Springer
Partially Observable Markov Decision Processes (POMDP) provide a standard framework for sequential decision making in stochastic environments. In this setting, an agent takes actio...
97
Voted
CVPR
1999
IEEE
16 years 7 days ago
1999
IEEE
We present a novel mixed-state dynamic Bayesian network (DBN) framework for modeling and classifying timeseries data such as object trajectories. A hidden Markov model (HMM) of di...
105
click to vote
AVSS
2008
IEEE
15 years 4 months ago
2008
IEEE
We propose a video event analysis framework based on object segmentation and tracking, combined with a Hidden Semi-Markov Model (HSMM) that uses state occupancy duration modeling....