Jeff A. Bilmes

A High-Speed, Low-Resource ASR Back-End Based on Custom Arithmetic (2008)

Xiao Li, Jonathan Malkin, Jeff A. Bilmes

Abstract—With the skyrocketing popularity of mobile devices, new processing methods tailored to a specific application have become necessary for low-resource systems. This work presents a...

General Terms (2008)

Susumu Harada, James A. L, Jonathan Malkin, Xiao Li, Jeff A. Bilmes

Mouse control has become a crucial aspect of many modern day computer interactions. This poses a challenge for individuals with motor impairments or those whose use of hands are restricted due to...

Non-Minimal Triangulations for Mixed Stochastic/Deterministic Graphical Models (2008)

Chris D. Bartels, Jeff A. Bilmes

We observe that certain large-clique graph triangulations can be useful for reducing computational requirements when making queries on mixed stochastic/deterministic graphical models. We demonstrate...

WHAT HMMS CAN’T DO (2008)

Jeff A. Bilmes

Hidden Markov models (HMMs) are the predominant methodology for automatic speech recognition (ASR) systems. Ever since their inception, it has been said that HMMs are an inadequate statistical model...

Non-Minimal Triangulations for Mixed Stochastic/Deterministic Graphical Models (2008)

Chris D. Bartels, Jeff A. Bilmes

We observe that certain large-clique graph triangulations can be useful for reducing computational requirements when making queries on mixed stochastic/deterministic graphical models. We demonstrate...

General Terms (2008)

Susumu Harada, James A. L, Jonathan Malkin, Xiao Li, Jeff A. Bilmes

Mouse control has become a crucial aspect of many modern day computer interactions. This poses a challenge for individuals with motor impairments or those whose use of hands are restricted due to...

WHAT HMMS CAN’T DO (2008)

Jeff A. Bilmes

Hidden Markov models (HMMs) are the predominant methodology for automatic speech recognition (ASR) systems. Ever since their inception, it has been said that HMMs are an inadequate statistical model...

Discriminatively Structured Graphical Models for Speech Recognition The Graphical Models Team (2008)

Jhu Summer Workshop, Jeff A. Bilmes, Geoff Zweig Ibm, Karen Livescu Mit, Peng Xu, Kirk Jackson Dod, ...

In recent years there has been growing interest in discriminative parameter training techniques, resulting from notable improvements in speech recognition performance on tasks ranging in size from...

Modeling peptide fragmentation with dynamic Bayesian networks for peptide identification (2008)

Klammer, Aaron A., Reynolds, Sheila M., Bilmes, Jeff A., MacCoss, Michael J., Noble, William Stafford

Motivation: Tandem mass spectrometry (MS/MS) is an indispensable technology for identification of proteins from complex mixtures. Proteins are digested to peptides that are then identified by their...

1 USING MUTUAL INFORMATION TO DESIGN FEATURE COMBINATIONS (2007)

Jeff A. Bilmes

Combination of different feature streams is a well-established method for improving speech recognition performance. This empirical success, however, poses theoretical problems when trying to design...

Multi-dynamic Bayesian networks (2006)

Karim Filali, Jeff A. Bilmes

We present a generalization of dynamic Bayesian networks to concisely describe complex probability distributions such as in problems with multiple interacting variable-length streams of random...

The Vocal Joystick data collection effort and vowel corpus (2006)

Kelley Kilanski, Jonathan Malkin, Xiao Li, Richard Wright, Jeff A. Bilmes

Vocal Joystick is a mechanism that enables individuals with motor impairments to make use of vocal parameters to control objects on a computer screen (buttons, sliders, etc.) and ultimately will be...

and et.al., “The Vocal Joystick (2006)

Jeff A. Bilmes, Jonathan Malkin, Xiao Li, Susumu Harada, Kelley Kilanski, Katrin Kirchhoff, ...

The Vocal Joystick is a novel human-computer interface mechanism designed to enable individuals with motor impairments to make use of vocal parameters to control objects on a computer screen...

What HMMs Can Do (2006)

BILMES, Jeff A.

Since their inception almost fifty years ago, hidden Markov models (HMMs) have have become the predominant methodology for automatic speech recognition (ASR) systems—today, most state-of-the-art...

☞ SVMs: Maximize the margin (2005)

Jeff A. Bilmes, Patrick Haffner, Nd Half Kernel, Margin Classifiers

➠ Good classifiers for speech and language applications? ➠ Huge range of approaches: focus on recent advances. ☞ Stronger emphasis on natural language applications: received a lot of recent...

The Vocal Joystick: A Voice-Based Human-Computer Interface for Individuals with Motor Impairments (2005)

Jeff A. Bilmes, Xiao Li, Jonathan Malkin, Kelley Kilanski, Richard Wright, Katrin Kirchhoff, ...

We present a novel voice-based humancomputer interface designed to enable individuals with motor impairments to use vocal parameters for continuous control tasks. Since discrete spoken commands are...

The Vocal Joystick Demo at UIST05: A Voice-Based Human-Computer Interface (2005)

Jeff A. Bilmes, Xiao Li, Jonathan Malkin, Kelley Kilanski, Richard Wright, Amarnag Subramanya, ...

We will demonstrate a novel voice-based human-computer interface we call the Vocal Joystick (VJ), designed to enable individuals with motor impairments to use vocal parameters for both discrete and...

The Vocal Joystick: A Voice-Based Human-Computer Interface for Individuals with Motor Impairments (2005)

Jeff A. Bilmes, Xiao Li, Jonathan Malkin, Kelley Kilanski, Richard Wright, Katrin Kirchhoff, ...

We present a novel voice-based humancomputer interface designed to enable individuals with motor impairments to use vocal parameters for continuous control tasks. Since discrete spoken commands are...

The Vocal Joystick: A Voice-Based Human-Computer Interface for Individuals with Motor Impairments (2005)

Jeff A. Bilmes, Xiao Li, Jonathan Malkin, Kelley Kilanski, Richard Wright, Katrin Kirchhoff, ...

We present a novel voice-based humancomputer interface designed to enable individuals with motor impairments to use vocal parameters for continuous control tasks. Since discrete spoken commands are...

The Vocal Joystick: A Voice-Based Human-Computer Interface for Individuals with Motor Impairments (2005)

Jeff A. Bilmes, Xiao Li, Jonathan Malkin, Kelley Kilanski, Richard Wright, Amarnag Subramanya, ...

We describe the technical details behind a novel voicebased human-computer interface designed to enable individuals with motor impairments to use vocal parameters for both discrete and continuous...

Part-of-speech tagging using virtual evidence and negative training (2005)

Sheila M. Reynolds, Jeff A. Bilmes

We present a part-of-speech tagger which introduces two new concepts: virtual evidence in the form of an “observed child” node, and negative training data to learn the conditional probabilities...

A generative/discriminative learning algorithm for image classification (2005)

Yi Li, Linda G. Shapiro, Jeff A. Bilmes

We have developed a two-phase generative / discriminative learning procedure for the recognition of classes of objects and concepts in outdoor scenes. Our method uses both multiple types of object...

Bilmes. Statistical models for empirical search-based performance tuning (2004)

Richard Vuduc, Richard Vuduc, James W. Demmel, James W. Demmel, Jeff A. Bilmes, Jeff A. Bilmes

Achieving peak performance from the computational kernels that dominate application performance often requires extensive machine-dependent tuning by hand. Automatic tuning systems have emerged in...

References for: Graphical Model Research in Audio, Speech, and Language Processing (2003)

Jeff A. Bilmes

This document provides a list of references to accompany the tutorial entitled as above and presented during the 2003 Uncertainty in Artificial Intelligence (UAI’03) conference, and mentioned at...

On triangulating dynamic graphical models (2003)

Jeff A. Bilmes, Chris Bartels

This paper introduces improved methodology to triangulate dynamic graphical models and dynamic Bayesian networks (DBNs). In this approach, a standard DBN template can be modified so the repeating and...

Factored language models and generalized parallel backoff (2003)

Jeff A. Bilmes, Katrin Kirchhoff

We introduce factored language models (FLMs) and generalized parallel backoff (GPB). An FLM represents words as bundles of features (e.g., morphological classes, stems, data-driven clusters, etc.),...

Factored Language Models and Generalized Parallel Backoff (2003)

Jeff A. Bilmes, Katrin Kirchhoff

We introduce factored language models (FLMs) and generalized parallel backoff (GPB). An FLM represents words as bundles of features (e.g., morphological classes, stems, data-driven clusters, etc.),...

Data-Driven Vector Clustering for Low-Memory Footprint ASR (2002)

Karim Filali, Li Xiao, Jeff A. Bilmes

It is important to produce automatic speech recognition (ASR) systems that use as few computational and memory resources as possible, especially in low-memory/low-power environments such as for...

Frontend PostProcessing and Backend Model Enhancement on (2002)

Chia-ping Chen, Karim Filali, Jeff A. Bilmes

We investigate a highly effective and extremely simple noiserobust front end based on novel post-processing of standard MFCC features on the Aurora databases. It performs remarkably well on both the...

J.Bilmes, “Data-driven vector clustering for low-memory footprint ASR (2002)

Karim Filali, Li Xiao, Jeff A. Bilmes

It is important to produce automatic speech recognition (ASR) systems that use as few computational and memory resources as possible, especially in low-memory/low-power environments such as for...

Frontend Post-Processing And Backend Model Enhancement On The Aurora 2.0/3.0 Databases (2002)

Chia-ping Chen, Karim Filali, Jeff A. Bilmes

We investigate a highly effective and extremely simple noiserobust front end based on novel post-processing of standard MFCC features on the Aurora databases. It performs remarkably well on both the...

Factored sparse inverse covariance matrices (2000)

Jeff A. Bilmes

Most HMM-based speech recognition systems use Gaussian mixtures as observation probability density functions. An important goal in all such systems is to improve parsimony. One method is to adjust...

Combination and joint training of acoustic classifiers for speech recognition (2000)

Katrin Kirchhoff, Jeff A. Bilmes

Classifier combination is a technique that often provides significant improvements in accuracy, and also furnishes a useful mechanism to support multi-modal information sources. In this paper we...

Directed graphical models of classifier combination: Application to phone recognition (2000)

Jeff A. Bilmes, Katrin Kirchhoff

Classifier combination is a technique that often provides appreciable accuracy gains. In this paper, we argue that the underlying statistical model of classifier combination should be made explicit....

Using Mutual Information To Design Feature Combinations (2000)

Jeff A. Bilmes

Combination of different feature streams is a well-established method for improving speech recognition performance. This empirical success, however, poses theoretical problems when trying to design...

Directed graphical models of classifier combination: Application to phone recognition (2000)

Jeff A. Bilmes, Katrin Kirchhoff

Classifier combination is a technique that often provides appreciable accuracy gains. In this paper, we argue that the underlying statistical model of classifier combination should be made explicit....

Factored sparse inverse covariance matrices (2000)

Jeff A. Bilmes

Most HMM-based speech recognition systems use Gaussian mixtures as observation probability density functions. An important goal in all such systems is to improve parsimony. One method is to adjust...

Factored sparse inverse covariance matrices (2000)

Jeff A. Bilmes

Most HMM-based speech recognition systems use Gaussian mixtures as observation probability density functions. An important goal in all such systems is to improve parsimony. One method is to adjust...

Statistical acoustic indications of coarticulation (1999)

Katrin Kirchhoff, Jeff A. Bilmes

Coarticulation in speech is one of the most difficult problems for automatic speech recognition (ASR) systems. The degree of coarticulation is assumed to vary with contextual conditions, such as...

Dynamic classifier combination in hybrid speech recognition systems using utterance-level confidence values (1999)

Katrin Kirchhoff, Ag Angewandte Informatik, Jeff A. Bilmes

A recent development in the hybrid HMM/ANN speech recognition paradigm is the use of several subword classifiers, each of which provides different information about the speech signal. Although the...

Buried Markov Models For Speech Recognition (1999)

Jeff A. Bilmes

Good HMM-based speech recognition performance requires at most minimal inaccuracies to be introduced by HMM conditional independence assumptions. In this work, HMM conditional independence...

Statistical Acoustic Indications Of Coarticulation (1999)

Katrin Kirchhoff, Jeff A. Bilmes

Coarticulation in speech is one of the most difficult problems for automatic speech recognition (ASR) systems. The degree of coarticulation is assumed to vary with contextual conditions, such as...

Dynamic Classifier Combination In Hybrid Speech Recognition Systems Using Utterance-Level Confidence Values (1999)

Katrin Kirchhoff, Ag Angewandte Informatik, Jeff A. Bilmes

A recent development in the hybrid HMM/ANN speech recognition paradigm is the use of several subword classifiers, each of which provides different information about the speech signal. Although the...

Statistical Acoustic Indications Of Coarticulation (1999)

Katrin Kirchhoff, Jeff A. Bilmes

Coarticulation in speech is one of the most difficult problems for automatic speech recognition (ASR) systems. The degree of coarticulation is assumed to vary with contextual conditions, such as...

Buried Markov Models for Speech Recognition (1999)

Jeff A. Bilmes

Good HMM-based speech recognition performance requires at most minimal inaccuracies to be introduced by HMM conditional independence assumptions. In this work, HMM conditional independence...

Data-Driven Extensions To HMM Statistical Dependencies (1998)

Jeff A. Bilmes

In this paper, a new technique is introduced that relaxes the HMM conditional independence assumption in a principled way. Without increasing the number of states, the modeling power of an HMM is...

Maximum Mutual Information Based Reduction Strategies for Cross-Correlation Based Joint Distributional Modeling (1998)

Jeff A. Bilmes

In maximum-likelihood based speech recognition systems, it is important to accurately estimate the joint distribution of feature vectors given a particular acoustic model. In previous work, we showed...

A Gentle Tutorial of the EM Algorithm and its Application to Parameter Estimation for Gaussian Mixture and Hidden Markov Models (1998)

Jeff A. Bilmes

We describe the maximum-likelihood parameter estimation problem and how the ExpectationMaximization (EM) algorithm can be used for its solution. We first describe the abstract form of the EM...

Estimation for Gaussian Mixture and Hidden Markov Models (1998)

Jeff A. Bilmes

We describe the maximum-likelihood parameter estimation problem and how the Expectation-form of the EM algorithm as it is often given in the literature. We then develop the EM parameter estimation...

A gentle tutorial on the EM algorithm and its application to parameter estimation for gaussian mixture and hidden markov models (1997)

Jeff A. Bilmes

We describe the maximum-likelihood parameter estimation problem and how the Expectation-form of the EM algorithm as it is often given in the literature. We then develop the EM parameter estimation...

Joint distributional modeling with cross-correlation based features (1997)

Jeff A. Bilmes

Abstract- In maximum-likelihood based speech recognition systems, it is important to accurately estimate the joint distribution of feature vectors given a particular acoustic model. In this work, we...

A gentle tutorial on the EM algorithm and its application to parameter estimation for gaussian mixture and hidden markov models (1997)

Jeff A. Bilmes

We describe the maximum-likelihood parameter estimation problem and how the Expectation-form of the EM algorithm as it is often given in the literature. We then develop the EM parameter estimation...

A gentle tutorial on the EM algorithm and its application to parameter estimation for gaussian mixture and hidden markov models (1997)

Jeff A. Bilmes

We describe the maximum-likelihood parameter estimation problem and how the Expectation-form of the EM algorithm as it is often given in the literature. We then develop the EM parameter estimation...

Joint Distributional Modeling with Cross-Correlation Based Features (1997)

Jeff A. Bilmes

In maximum-likelihood based speech recognition systems, it is important to accurately estimate the joint distribution of feature vectors given a particular acoustic model. In this work, we propose...

Joint distributional modeling with cross-correlation based features (1997)

Jeff A. Bilmes

Abstract- In maximum-likelihood based speech recognition systems, it is important to accurately estimate the joint distribution of feature vectors given a particular acoustic model. In this work, we...

Joint Distributional Modeling with Cross-Correlation. Based Features (1997)

Jeff A. Bilmes

In maximum-likelihood based speech recognition systems, it is important to accurately estimate the joint distribution of feature vectors given a particular acoustic model. In this work, we propose...

Transmembrane Topology and Signal Peptide Prediction Using Dynamic Bayesian Networks

Reynolds, Sheila M., Käll, Lukas, Riffle, Michael E., Bilmes, Jeff A., Noble, William Stafford

Hidden Markov models (HMMs) have been successfully applied to the tasks of transmembrane protein topology prediction and signal peptide prediction. In this paper we expand upon this work by making...

Modeling peptide fragmentation with dynamic Bayesian networks for peptide identification

Klammer, Aaron A., Reynolds, Sheila M., Bilmes, Jeff A., MacCoss, Michael J., Noble, William Stafford

Motivation: Tandem mass spectrometry (MS/MS) is an indispensable technology for identification of proteins from complex mixtures. Proteins are digested to peptides that are then identified by their...