D. Williams, "Knowing what you don't know: roles for confidence measures in automatic speech recognition," Ph.D. thesis, University of Sheffield, England, 1999. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Turkish LVCSR: Database Preparation and Language Modeling.. - Mengusoglu, Deroo (2001) (1 citation) (Correct)
....languages. They are promising to give better results for Turkish large vocabulary speech recognition. We have also described the collection of Turkish database. We are also planning to use the language modeling techniques studied in this paper, to improve the efficiency of confidence measures [13] [12] defined on acoustic models of speech. 6. ....
G. Williams, "Knowing What You Don't Know: Roles for Confidence Measures in Automatic Speech Recognition", PhD Thesis, Department of Computer Science, University of Sheffield, 1999.
Dynamic Pronunciation Models for Automatic Speech Recognition - Fosler-Lussier (1999) (3 citations) (Correct)
....this work. First and foremost, in order to fairly evaluate the dynamic pronunciation models in the next chapter, I wanted to build the best possible static dictionary, for use within the first pass recognizer. 1 This chapter contains some experiments previously reported in [Fosler Lussier and Williams, 1999] and [Fosler Lussier, 1999] 5.1. INTRODUCTION 77 Since dynamic dictionaries require a second decoding pass over lattices or n best lists of hypotheses generated by a first recognition pass, a good static dictionary was also necessary to generate the first pass word hypotheses. Another goal of ....
....training set. 94 CHAPTER 5. STATIC DICTIONARIES even if close in the acoustic score, contributes nothing to the final model. One way to describe the relative differences among models is to use an acoustic confidence score, which gives an estimate of model quality for a given segment of speech. Williams [1999] describes a technique for computing acoustic confidences for individual phones and complete words using a neural network acoustic model. The multi layer perceptron acoustic probability estimator used in the ICSI recognition system produces posterior probabilities of the form P (q n k jX ....
D. A. G. Williams. Knowing what you don't know: roles for confidence measures in automatic speech recognition. Ph.D. thesis, University of Sheffield, Sheffield, England, 1999.
On the Use of Quality Measures for.. - Garcia-Romero.. (2004) (Correct)
No context found.
D. Williams, "Knowing what you don't know: roles for confidence measures in automatic speech recognition," Ph.D. thesis, University of Sheffield, England, 1999.
Improved Pronunciation Modeling by Properly Integrating.. - Tsai, Chou, Lee (2002) (Correct)
No context found.
G.Williams, " Knowing What You Don't Know: Roles for Confidence Measures in Automatic Speech Recognition", Ph.D. Thesis, University of Sheffield, 1999.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC