Results 11  20
of
1,132
The Hierarchical Hidden Markov Model: Analysis and Applications
 MACHINE LEARNING
, 1998
"... . We introduce, analyze and demonstrate a recursive hierarchical generalization of the widely used hidden Markov models, which we name Hierarchical Hidden Markov Models (HHMM). Our model is motivated by the complex multiscale structure which appears in many natural sequences, particularly in langua ..."
Abstract

Cited by 326 (3 self)
 Add to MetaCart
. We introduce, analyze and demonstrate a recursive hierarchical generalization of the widely used hidden Markov models, which we name Hierarchical Hidden Markov Models (HHMM). Our model is motivated by the complex multiscale structure which appears in many natural sequences, particularly in language, handwriting and speech. We seek a systematic unsupervised approach to the modeling of such structures. By extendingthe standard forwardbackward(BaumWelch) algorithm, we derive an efficient procedure for estimating the model parameters from unlabeled data. We then use the trained model for automatic hierarchical parsing of observation sequences. We describe two applications of our model and its parameter estimation procedure. In the first application we show how to construct hierarchical models of natural English text. In these models different levels of the hierarchy correspond to structures on different length scales in the text. In the second application we demonstrate how HHMMs can b...
Tractable inference for complex stochastic processes
 In Proc. UAI
, 1998
"... The monitoring and control of any dynamic system depends crucially on the ability to reason about its current status and its future trajectory. In the case of a stochastic system, these tasks typically involve the use of a belief state—a probability distribution over the state of the process at a gi ..."
Abstract

Cited by 302 (14 self)
 Add to MetaCart
(Show Context)
The monitoring and control of any dynamic system depends crucially on the ability to reason about its current status and its future trajectory. In the case of a stochastic system, these tasks typically involve the use of a belief state—a probability distribution over the state of the process at a given point in time. Unfortunately, the state spaces of complex processes are very large, making an explicit representation of a belief state intractable. Even in dynamic Bayesian networks (DBNs), where the process itself can be represented compactly, the representation of the belief state is intractable. We investigate the idea of maintaining a compact approximation to the true belief state, and analyze the conditions under which the errors due to the approximations taken over the lifetime of the process do not accumulate to make our answers completely irrelevant. We show that the error in a belief state contracts exponentially as the process evolves. Thus, even with multiple approximations, the error in our process remains bounded indefinitely. We show how the additional structure of a DBN can be used to design our approximation scheme, improving its performance significantly. We demonstrate the applicability of our ideas in the context of a monitoring task, showing that orders of magnitude faster inference can be achieved with only a small degradation in accuracy. 1
Dialogue act modeling for automatic tagging and recognition of conversational speech
 COMPUTATIONAL LINGUISTICS
, 2000
"... We describe a statistical approach for modeling dialogue acts in conversational speech, i.e., speecactlike ..."
Abstract

Cited by 278 (14 self)
 Add to MetaCart
We describe a statistical approach for modeling dialogue acts in conversational speech, i.e., speecactlike
Operations for Learning with Graphical Models
 Journal of Artificial Intelligence Research
, 1994
"... This paper is a multidisciplinary review of empirical, statistical learning from a graphical model perspective. Wellknown examples of graphical models include Bayesian networks, directed graphs representing a Markov chain, and undirected networks representing a Markov field. These graphical models ..."
Abstract

Cited by 276 (13 self)
 Add to MetaCart
This paper is a multidisciplinary review of empirical, statistical learning from a graphical model perspective. Wellknown examples of graphical models include Bayesian networks, directed graphs representing a Markov chain, and undirected networks representing a Markov field. These graphical models are extended to model data analysis and empirical learning using the notation of plates. Graphical operations for simplifying and manipulating a problem are provided including decomposition, differentiation, and the manipulation of probability models from the exponential family. Two standard algorithm schemas for learning are reviewed in a graphical framework: Gibbs sampling and the expectation maximization algorithm. Using these operations and schemas, some popular algorithms can be synthesized from their graphical specification. This includes versions of linear regression, techniques for feedforward networks, and learning Gaussian and discrete Bayesian networks from data. The paper conclu...
Speaker recognition: A tutorial
"... A tutorial on the design and development of automatic speakerrecognition systems is presented. Automatic speaker recognition is the use of a machine to recognize a person from a spoken phrase. These systems can operate in two modes: to identify a particular person or to verify a person’s claimed id ..."
Abstract

Cited by 269 (2 self)
 Add to MetaCart
A tutorial on the design and development of automatic speakerrecognition systems is presented. Automatic speaker recognition is the use of a machine to recognize a person from a spoken phrase. These systems can operate in two modes: to identify a particular person or to verify a person’s claimed identity. Speech processing and the basic components of automatic speakerrecognition systems are shown and design tradeoffs are discussed. Then, a new automatic speakerrecognition system is given. This recognizer performs with 98.9 % correct identification. Last, the performances of various systems are compared.
RoboticsBased Location Sensing Using Wireless Ethernet
 Wireless Networks
, 2005
"... A key subproblem in the construction of locationaware systems is the determination of the position of a mobile device. This article describes the design, implementation and analysis of a system for determining position inside a building from measured RF signal strengths of packets on an IEEE 802.11 ..."
Abstract

Cited by 241 (5 self)
 Add to MetaCart
A key subproblem in the construction of locationaware systems is the determination of the position of a mobile device. This article describes the design, implementation and analysis of a system for determining position inside a building from measured RF signal strengths of packets on an IEEE 802.11b wireless Ethernet network. Previous approaches to locationawareness with RF signals have been severely hampered by nonGaussian signals, noise, and complex correlations due to multipath effects, interference and absorption. The design of our system begins with the observation that determining position from complex, noisy and nonGaussian signals is a wellstudied problem in the field of robotics. Using only offtheshelf hardware, we achieve robust position estimation to within a meter in our experimental context and after adequate training of our system. We can also coarsely determine our orientation and can track our position as we move. Our results show that we can localize a stationary device to within 1.5 meters over 80 % of the time and track a moving device to within 1 meter over 50 % of the time. Both localization and tracking run in realtime. By applying recent advances in probabilistic inference of position and sensor fusion from noisy signals, we show that the RF emissions from base stations as measured by offtheshelf wireless Ethernet cards are sufficiently rich in information to permit a mobile device to reliably track its location.
An Online Mapping Algorithm for Teams of Mobile Robots
 International Journal of Robotics Research
, 2001
"... We propose a new probabilistic algorithm for online mapping of unknown environments with teams of robots. At the core of the algorithm is a technique that combines fast maximum likelihood map growing with a Monte Carlo localizer that uses particle representations. The combination of both yields an o ..."
Abstract

Cited by 235 (14 self)
 Add to MetaCart
We propose a new probabilistic algorithm for online mapping of unknown environments with teams of robots. At the core of the algorithm is a technique that combines fast maximum likelihood map growing with a Monte Carlo localizer that uses particle representations. The combination of both yields an online algorithm that can cope with large odometric errors typically found when mapping an environment with cycles. The algorithm can be implemented distributedly on multiple robot platforms, enabling a team of robots to cooperatively generate a single map of their environment. Finally, an extension is described for acquiring threedimensional maps, which capture the structure and visual appearance of indoor environments in 3D.
The Power of Amnesia: Learning Probabilistic Automata with Variable Memory Length
 Machine Learning
, 1996
"... . We propose and analyze a distribution learning algorithm for variable memory length Markov processes. These processes can be described by a subclass of probabilistic finite automata which we name Probabilistic Suffix Automata (PSA). Though hardness results are known for learning distributions gene ..."
Abstract

Cited by 226 (17 self)
 Add to MetaCart
(Show Context)
. We propose and analyze a distribution learning algorithm for variable memory length Markov processes. These processes can be described by a subclass of probabilistic finite automata which we name Probabilistic Suffix Automata (PSA). Though hardness results are known for learning distributions generated by general probabilistic automata, we prove that the algorithm we present can efficiently learn distributions generated by PSAs. In particular, we show that for any target PSA, the KLdivergence between the distribution generated by the target and the distribution generated by the hypothesis the learning algorithm outputs, can be made small with high confidence in polynomial time and sample complexity. The learning algorithm is motivated by applications in humanmachine interaction. Here we present two applications of the algorithm. In the first one we apply the algorithm in order to construct a model of the English language, and use this model to correct corrupted text. In the second ...
An Efficient Probabilistic ContextFree Parsing Algorithm that Computes Prefix Probabilities
 Computational Linguistics
, 2002
"... this article can compute solutions to all four of these problems in a single flamework, with a number of additional advantages over previously presented isolated solutions ..."
Abstract

Cited by 222 (5 self)
 Add to MetaCart
(Show Context)
this article can compute solutions to all four of these problems in a single flamework, with a number of additional advantages over previously presented isolated solutions