Some approaches to statistical and finite-state speech-to-speech translation

Casacuberta, Francisco; Ney, Hermann; Och, Franz Josef; Vidal, Enrique; Vilar, Juan Miguel; Barrachina, Sergio; García-Varea, Ismael; Llorens, D.; Martínez, César; Molau, Sirko; Nevado, Francisco; Pastor, María Luisa Carrió; Picó, David; Sanchís, Alberto; Tillmann, Christoph

doi:10.1016/s0885-2308(03)00028-7

Cited by 63 publications

(32 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, recent advances in the statistical translation research field increase the chances of a widespread usage in the near future [1] [2]. One of the main issues in the deployment of statistical translators is the combination of different knowledge sources.…”

Section: Introductionmentioning

confidence: 99%

Statistical speech translation system based on voice recognition optimization using multimodal sources of knowledge and characteristics vectors

Cánovas

Tomás

Lloret

et al. 2013

Computer Standards & Interfaces

View full text Add to dashboard Cite

Abstract-Synergic combination of different sources of knowledge is a key issue for the development of modern statistical translators. Reconnaissance, and thus the translation, can be improved by adding new heuristic characteristics. In this work, a speech translation statistical system that adds additional other-than-voice information in a voice translation system is presented. The additional information serves as a base for the loglinear combination of several statistical models. We obtain the characteristics vectors using a statistical model that is based on the N-best reconnaissance list. We describe the theoretical framework of the problem, summarize the overall architecture of the system, and show how the system is enhanced with the additional information. Our real prototype implements a real-time speech translation system from Spanish to English that is adapted to specific teaching-related environments. Finally, we will provide and explain the system performance results. A tool like the one presented in this article may increase the participation rate of the foreign students to the lecture classes and talks.

show abstract

Section: Introductionmentioning

confidence: 99%

Statistical speech translation system based on voice recognition optimization using multimodal sources of knowledge and characteristics vectors

Cánovas

Tomás

Lloret

et al. 2013

Computer Standards & Interfaces

View full text Add to dashboard Cite

show abstract

“…Statistical machine translation (SMT) methods have evolved from using the simple word based models to phrase based models (Marcu and Wong, 2002;Och and Ney, 2004). More recently, there is a significant effort focusing on integrating richer knowledge, such as syntactic parse trees (Huang and Knight, 2006) within the translation process to overcome the limitations of the phrase based models.…”

Section: Introductionmentioning

confidence: 99%

Automatic acquisition of grammatical types for nouns

Bel

Espeja

Marimon

2007

Human Language Technologies 2007: The Conference of the North American Chapter of the Association for Computational Linguistics

View full text Add to dashboard Cite

The work 1 we present here is concerned with the acquisition of deep grammatical information for nouns in Spanish. The aim is to build a learner that can handle noise, but, more interestingly, that is able to overcome the problem of sparse data, especially important in the case of nouns. We have based our work on two main points. Firstly, we have used distributional evidences as features. Secondly, we made the learner deal with all occurrences of a word as a single complex unit. The obtained results show that grammatical features of nouns is a level of generalization that can be successfully approached with a Decision Tree learner.

show abstract

“…The WER metric [Amengual et al, 2000a, Casacuberta et al, 2004 is defined as the minimum number of word substitution, deletion and insertion operations required to convert the target sentence provided by the translation system into the reference translation, divided by the number of words of the reference translation. It can also be seen as the ratio of the edit distance between the system and the reference translation, and the number of words of the reference translation.…”

Section: Automatic Mt Evaluation Metricsmentioning

confidence: 99%

“…Even though these models are much more limited than other more powerful ones, the computational costs of the algorithms that are needed to deal with them are much lower. SFSTs also permit a simple integration with other information sources, which makes it easy to apply SFSTs to more difficult tasks such as speech translation [Casacuberta et al, 2004]. SFSTs and the corresponding training and search techniques have been studied by several authors, in many cases explicitly motivated by MT applications [E. Vidal and Segarra, 1989, Oncina et al, 1993, Knight and Al-Onaizan, 1998, Mäkinen, 1999, Amengual et al, 2000, Alshawi et al, 2000a, Casacuberta, 2000a, Vilar, 2000, Vogel and Ney, 2000, Picó and Casacuberta, 2001, Bangalore and Riccardi, 2003, Kumar and Byrne, 2003, Casacuberta and Vidal, 2004, Tsukada and Nagata, 2004, Casacuberta et al, 2005, Kumar et al, 2006, Casacuberta and Vidal, 2007, Mariòo et al, 2006.…”

Section: Introductionmentioning

confidence: 99%

“…The GIATI technique [Casacuberta and Vidal, 2007] has been applied to machine translation [Casacuberta and Vidal, 2004], speech translation [Casacuberta et al, 2004] and computed-assisted translation [Barrachina et al]. The results obtained using GIATI suggest that, among all the SFST learning techniques tested, GIATI is the only one that can cope with translation tasks under real conditions of vocabulary sizes and amounts of training data available.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Statistical approaches for natural language modelling and monotone statistical machine translation

Ferrer¹

View full text Add to dashboard Cite

AcknowledgementsEstas líneas son las más difíciles de escribir de la tesis aunque no contengan formulas ni formalismos matemáticos. Si tuviese que enumerar a todas las personas que han afectado positivamente (ó negativamente) al resultado de esta tesis, probablemente, tendría que anotar toda la tesis con comentarios al margen. Por ello, solo nombraré a las personas cuyas aportaciones han sido constantes e/o importantes sin por ello menospreciar al resto de personas que han influido.Comenzaré por proximidad, agradeciendo a los habitantes del laboratirio 101L del departamento. No sólo a los miembros actuales, sino también a los fundadores de "Teachings of the final convergence". En particular, agradecer a Ramón Granell por nuestras discusiones; a Adrián Gimenez por ejercer de "diablo" en las reflexiones de algunos de mis problemas; y a Ricardo Sánchez por ser tan "coloquial y amigable".Esta tésis no habría existido de no ser por el apoyo económico de la "Generalitat Valenciana" formalizado en la beca FPI con referencia CTBPRA/2005/004 recibida bajo el amparo de la "Conselleria d'Empresa, Universitat i Ciència". Así mismo estoy en deuda con Alfons Juan Císcar tanto por haber aceptado ser mi director de tesis, como por su consejo y su tesón en evitar mi dispersión en tantos temas interesantes que existen en la investigación. También debo agradecimiento a Francisco Casacuberta Nolla por su reflexiones, correcciones y guía. A parte del agradecimiento a mis directores de tesis me gustaría agradecerle a Prof. Hermann Ney por haberme acogido en Aquisgrán permitiendome así "disfrutar" tanto del clima de dicha región germana, como de sus apasionantes reflexiones que me permitieron refinar e incrementar mis conocimientos.Mención especial requieren todos de mis compañeros de fatigas del PRHLT y del ITI en general y algunos en particular como por ejemplo Jorge Civera por soportar mis reflexiones sobre asuntos no sólo de investigación y por la plantilla de esta tesis que me ha ahorrado muchos quebraderos de cabeza; o Daniel Ortiz e Ismael García Varea por nuestras colaboraciones que son parte de esta tesis; así como a todos aquellos que con asiduidad han asistido las cenas minimalistas y no minimalistas.Apropiandome de una cita del célebre Adrián Giménez "Cada tesis esconde un drama personal"; y este caso no podría ser una excepción. Así que estoy en deuda con todos aquellos que me han permitido disfrutar de la vida durante esta ardua e interminable tarea. Me veo en el deber de comenzar por la fuerte amistad que forgé con mis compañeros de carrera: Alex, Gabi y Jesús; amistad que todavía hoy perdura. v Desde mi retiro como exremero, no puedo sino que agradecer a todo el equipo de remo de la Universidad Politécnica de Valencia, tanto por las millas y el sudor derramado en la mar; como por las cervezas ingeridas y derramadas en la mesa de algún bar. En especial a todos aquellos remeros que me han acompañado desde el comienzo no sólo de esta tesis sino de mis estudios en informática cuando decicí practicar tan gratificante deport...

show abstract

Some approaches to statistical and finite-state speech-to-speech translation

Cited by 63 publications

References 30 publications

Statistical speech translation system based on voice recognition optimization using multimodal sources of knowledge and characteristics vectors

Statistical speech translation system based on voice recognition optimization using multimodal sources of knowledge and characteristics vectors

Automatic acquisition of grammatical types for nouns

Statistical approaches for natural language modelling and monotone statistical machine translation

Contact Info

Product

Resources

About