Thomas Niesler, Daniel Willett
We perform language identification experiments for four prominent South-African languages using a multilingual speech recognition system. Specifically, we show how successfully Afrikaans, English,...
PARAPHRASING SPONTANEOUS SPEECH USING WEIGHTED FINITE-STATE TRANSDUCERS (2008)
Takaaki Hori, Daniel Willett, Yasuhiro Minami
This paper describes an integrated framework to paraphrase spontaneous speech into written-style sentences. Most current speech recognition systems try to transcribe whole spoken words correctly....
Daniel Willett, Gerhard Rigoll
. In this paper, we focus on a novel NN/HMM architecture for continuous speech recognition. The architecture incorporates a neural feature extraction to gain more discriminative feature vectors for...
REDUCED LEXICON TREES FOR DECODING IN A MMI-CONNECTIONIST/HMM SPEECH RECOGNITION SYSTEM (2007)
Christoph Neukirchen, Daniel Willett, Gerhard Rigoll
The presented work deals with the experimental identification of parts in a tree based decoder lexicon, that are more important for decoding efficiency compared to less important lexicon parts. Three...
Daniel Willett, Erik Mcdermott, Yasuhiro Minami, Atsushi Nakamura, Shigeru Katagiri
We present our efforts to set up an efficient timesynchronous Viterbi-decoder for arbitrary speech recognition tasks. This decoder adopts the recently proposed framework of representing the...
A DISCRIMINATIVE TRAINING PROCEDURE BASED ON LANGUAGE MODEL AND DICTIONARY FOR LVCSR (2007)
Daniel Willett, Stefan Muller, Gerhard Rigoll
In today's HMM-based speech recognition systems, the parameters are most commonly estimated according to the Maximum Likelihood criterion. Because of limited training data, however,...
UNSUPERVISED PRONUNCIATION ADAPTATION FOR OFF-LINE TRANSCRIPTION OF JAPANESE LECTURE SPEECHES (2007)
Daniel Willett, Erik Mcdermott, Shigeru Katagiri
Observing that most variations in pronunciation are strongly speaker and speaking style dependent, and that the introduction of pronunciation variants in a speaker-independent recognition system is...
The thesis deals with different aspects of automatic speech recognition. After an introduction, which describes the most important fundamental ideas, methodologies and algorithms, some new approaches...
The thesis deals with different aspects of automatic speech recognition. After an introduction, which describes the most important fundamental ideas, methodologies and algorithms, some new approaches...
The thesis deals with different aspects of automatic speech recognition. After an introduction, which describes the most important fundamental ideas, methodologies and algorithms, some new approaches...
The thesis deals with different aspects of automatic speech recognition. After an introduction, which describes the most important fundamental ideas, methodologies and algorithms, some new approaches...
A continuous density interpretation of discrete HMM systems and MMI-neural networks (2001)
Christoph Neukirchen, Jörg Rottl, Daniel Willett, Gerhard Rigoll, Senior Member
Abstract—The subject of this paper is the integration of the traditional vector quantizer (VQ) and discrete hidden Markov models (HMM) combination in the mixture emission density framework commonly...
Martha Larson, Daniel Willett, Joachim Köhler, Gerhard Rigoll
This paper proposes a novel combined compound splitting and phrase recombination method that optimizes the composition of the speech recognition lexicon for a given domain. Data-driven compound word...
Unlimited Vocabulary Script Recognition Using Character N-Grams (2000)
Anja Brakensiek, Daniel Willett, Gerhard Rigoll
. In this paper a robust script recognition system is described, which makes use of a language model, that consists of backo character n-grams. The system is based on Hidden Markov Models (HMMs)...
Improved Degraded Document Recognition with Hybrid Modeling Techniques and Character N-Grams (2000)
Anja Brakensiek, Daniel Willett, Gerhard Rigoll
In this paper a robust multifont character recognition system for degraded documents such as photocopy or fax is described. The system is based on Hidden Markov Models (HMMs) using discrete and...
Frame-Discriminative And Confidence-Driven Adaptation For LVCSR (2000)
Frank Wallhoff, Daniel Willett, Gerhard Rigoll
Maximum Likelihood Linear Regression (MLLR) has become the most popular approach for adapting speakerindependent Hidden Markov Models to a specic speaker's characteristics. However, it is well...
Ducoder - The Duisburg University Lvcsr Stackdecoder (2000)
Daniel Willett, Christoph Neukirchen, Gerhard Rigoll
With this paper, we present the DUcoder, the LVCSR decoder developed at Duisburg University. The decoder performs the Viterbi search for the most probable word sequence in recognition systems that...
Experiments In Topic Indexing Of Broadcast News Using Neural Networks (1999)
Christoph Neukirchen, Daniel Willett, Gerhard Rigoll
The paper deals with the problem of extracting topic information from news show stories by statistical methods. It is shown that the traditional topic-dependent n-gram language modeling approach can...
Jörg Rottland, J Org Rottl, Christoph Neukirchen, Daniel Willett, Gerhard Rigoll
This paper describes, how to perform speaker adaptation for a hybrid large vocabulary speech recognition system. The hybrid system is based on a Maximum Mutual Information Neural Network (MMINN),...
Daniel Willett, Christoph Neukirchen, Jörg Rottland, J Org Rottl, Gerhard Rigoll
Decision tree-based state clustering has emerged in recent years as the most popular approach for clustering the states of context dependent hidden Markov model based speech recognizers. The...
Confidence measures for hmm-based speech recognition (1998)
Daniel Willett, Andreas Worm, Christoph Neukirchen, Gerhard Rigoll
In this paper, we describe our work on the field of confidence measures for HMM-based speech recognition. Confidence measures are a means of estimating the recognition reliability for single words of...
Christoph Neukirchen, Daniel Willett, Stefan Eickeler, Stefan Müller
In previous work about hybrid speech recognizers with discrete HMMs we have shown that VQs, that are trained according to an MMI criterion, are well suited for ML estimated Bayes classifiers. This is...
Gerhard Rigoll, Daniel Willett
This paper deals with a hybrid NN/HMM architecture for continuous speech recognition. We present a novel approach to set up a neural linear or nonlinear feature transformation that is used as a...
Daniel Willett, Arno Rorner, Arno Romer, Jorg Rottland, Gerhard Rigoll
In this paper, we present the basic design principles and architecture of a dialogue system for scheduling appointments. This mixed-initiative dialogue system integrates an automatic...
Soft State-Tying For HMM-Based Speech Recognition (1998)
Christoph Neukirchen, Daniel Willett, Gerhard Rigoll
This paper introduces a method for regularization of HMM systems that avoids parameter overfitting causedby insufficient training data. Regularization is done by augmenting the EM training method by...
Efficient Search With Posterior Probability Estimates In Hmm-Based Speech Recognition (1998)
Daniel Willett, Christoph Neukirchen, Gerhard Rigoll
In this paper we present the methods we developed to estimate posterior probabilities for HMM states in continuous and discrete HMM-based speech recognition systems and several ways to speed up...
Daniel Willett, Christoph Neukirchen, Jörg Rottland
The estimation of the HMM parameters has always been a major issue in the design of speech recognition systems. Discriminative objectives like Maximum Mutual Information (MMI) or Minimum...
Hybrid NN/HMM-Based Speech Recognition with a Discriminant Neural Feature Extraction (1997)
Daniel Willett, Gerhard Rigoll
In this paper, we present a novel hybrid architecture for continuous speech recognition systems. It consists of a continuous HMM system extended by an arbitrary neural network that is used as a...
A New Approach To Generalized Mixture Tying For Continuous HMM-Based Speech Recognition (1997)
Daniel Willett, Gerhard Rigoll
In this paper we present a new approach for a generalized tying of mixture components for continuous mixture-density HMM-based speech recognition systems. With an iterative pruning and splitting...
Hybrid NN/HMM-Based Speech Recognition with a Discriminant Neural Feature Extraction (1997)
Daniel Willett, Gerhard Rigoll
In this paper, we present a novel hybrid architecture for continuous speech recognition systems. It consists of a continuous HMM system extended by an arbitrary neural network that is used as a...
Rapid Vector Quantization and Classification with Neural Networks (1996)
. The majority of today's Neural Networks are either MultiLayer -Perceptron networks (MLP) or Feature-Maps like Kohonen's SelfOrganizing Map (SOM). Usually they are simulated on ordinary...
Unsupervised Language Model Adaptation for Lecture Speech Transcription
Thomas Niesler, Daniel Willett
Unsupervised adaptation methods have been applied successfully to the acoustic models of speech recognition systems for some time. Relatively little work has been carried out in the area of...