In the last decade of automatic speech recognition (ASR) research, the introduction of deep
learning has brought considerable reductions in word error rate of more than 50% relative …

The Application of Hidden Markov Models in Speech Recognition Page 1 The Application of
Hidden Markov Models in Speech Recognition Full text available at: http://dx.doi.org/10.1561/2000000004 …

In this paper we introduce the Minimum Phone Error (MPE) and Minimum Word Error (MWE)
criteria for the discriminative training of HMM systems. The MPE/MWE criteria are smoothed …

A complete overview of distant automatic speech recognition The performance of
conventional Automatic Speech Recognition (ASR) systems degrades dramatically as soon …

This thesis investigates the use of discriminative criteria for training HMM parameters for
speech recognition, in particular the Maximum Mutual Information (MMI) criterion and a new …

This paper describes, and evaluates on a large scale, the lattice based framework for
discriminative training of large vocabulary speech recognition systems based on Gaussian …

Understanding architectural choices for deep neural networks (DNNs) is crucial to improving
state-of-the-art speech recognition systems. We investigate which aspects of DNN acoustic …

Conditional random fields (Lafferty et al., 2001) are quite effective at sequence labeling
tasks like shallow parsing (Sha and Pereira, 2003) and namedentity extraction (McCallum …

This paper examines how people communicate with computers using speech. Automatic
speech recognition (ASR) transforms speech into text, while automatic speech synthesis [or …

Bandwidth extension (BWE) refers to various methods that increase either the perceived or
real frequency spectrum (bandwidth) of audio signals. Such frequency extension is …