Moisés Pastor

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

Abstract Speech-input translation can be properly approached as a pattern recognition problem by means of statistical alignment models and stochastic finite-state transducers. Under this general framework, some specific models are presented. One of the features of such models is their capability of automatically learning from training examples. Moreover, the stochastic finite-state transducers permit an integrated architecture similar to one used in speech recognition. In this case, the acoustic models (hidden Markov models) are embedded into the finite-state transducers, and the translation of a source utterance is the result of a (Viterbi) search on the integrated network. These approaches have been followed in the framework of the European project E u T rans . Translation experiments have been performed from Spanish to English and from Italian to English in an application involving the interaction of a customer with a receptionist at the frontdesk of a hotel.

The EuTrans Spoken Language Translation System

The EuTransAll project aims at using example-based approaches for the automatic development of Machine Translation systems accepting text and speech input for limited-domain applications. During the first phase of the project, a speech-translation system that is based on the use of automatically learned subsequential transducers has been built. This paper contains a detailed and mostly self-contained overview of the transducer-learning algorithms and system architecture, along with a new approach for using categories representing words or short phrases in both input and output languages. Experimental results using this approach are reported for a task involving the recognition and translation of sentences in the hotel-receptioncommunication domain, with a vocabulary of 683 words in Spanish. Atranslation word-error rate of 1.97% is achieved in real-timefactor 2.7 on a Personal Computer.

Speech-to-speech translation based on finite-state transducers

Nowadays, the most successful speech recognition systems are based on stochastic finite-state networks (hidden Markov models and n-grams). Speech translation can be accomplished in a similar way as speech recognition. Stochastic finite-state transducers, which are specific stochastic finite-state networks, have proved very adequate for translation modeling. In this work a speech-to-speech translation system, the EuTRANS system, is presented. The acoustic, language and translation models are finite-state networks that are automatically learnt from training samples. This system was assessed in a series of translation experiments from Spanish to English and from Italian to English in an application involving the interaction (by telephone) of a customer with a receptionist at the front-desk of a hotel.

Projection Profile Based Algorithm for Slant Removal

The slant is one of the main sources of handwritten text variability. The slant is the clockwise angle between the vertical direction and the vertical text strokes. A well formalised and fast method to estimate the slant angle is presented. The method is based on the observation that the columns distribution of the vertical projection profile presents a maximum variance for the non slanted text. A comparative with Sobel operators convolution method and the Idiap slant method is provided.

Computer-assisted transcription of a historical botanical specimen book: organization and process overview

We describe a protocol designed for computer-assisted transcribing a XVII century botanical specimen book, based on Handwritten Text Recognition (HTR) technology. Here we focus on the organization and coordination aspects of this protocol and outline related technical issues. Using the proposed protocol, full ground truth data has been produced for the first book chapter and high-quality transcripts are being cost-effectively obtained for the rest of the approximately 1000 pages of the book. The process encompasses two main, computer-assisted steps; namely, image layout analysis and transcription. Layout analysis is based on a semi-supervised incremental approach and transcription makes use of an interactive-predictive HTR prototype known as CATTI. Currently, the first step of this procedure has been completed for the full book and the second step is close to be finished. Ultimately, all the data produced will be made publicly available for research and development.

Bi-modal handwritten text recognition (BiHTR) ICPR 2010 contest report

Handwritten text is generally captured through two main modalities: off-line and on-line. Each modality has advantages and disadvantages, but it seems clear that smart approaches to handwritten text recognition (HTR) should make use of both modalities in order to take advantage of the positive aspects of each one. A particularly interesting case where the need of this bi-modal processing arises is when an off-line text, written by some writer, is considered along with the online modality of the same text written by another writer. This happens, for example, in computer-assisted transcription of old documents, where on-line text can be used to interactively correct errors made by a main off-line HTR system. In order to develop adequate techniques to deal with this challenging bi-modal HTR recognition task, a suitable corpus is needed. We have collected such a corpus using data (word segments) from the publicly available off-line and on-line IAM data sets. In order to provide the Community with an useful corpus to make easy tests, and to establish baseline performance figures, we have proposed this handwritten bi-modal contest. Here is reported the results of the contest with two participants, one of them achieved a 0% classification error rate, whilst the other participant achieved an interesting 1.5%.

A Bi-modal Handwritten Text Corpus: Baseline Results

Handwritten text is generally captured through two main modalities: off-line and on-line. Smart approaches to handwritten text recognition (HTR) may take advantage of both modalities if they are available. This is for instance the case in computer-assisted transcription of text images, where on-line text can be used to interactively correct errors made by a main off-line HTR system. We present here baseline results on the biMod-IAM-PRHLT corpus, which was recently compiled for experimentation with techniques aimed at solving the proposed multi-modal HTR problem, and is being used in one of the official ICPR-2010 contests.

Baseline Detection on Arabic Handwritten Documents

Document processing comprises different steps depending on the nature of the documents. For text documents, specially for handwritten documents, transcription of their contents is one of the main tasks. Handwritten Text Recognition (HTR) is the process of automatically obtaining the transcription of the content of a handwritten text document. In document processing, the basic unit for the acquisition process is the page image, whilst line image is the basic form for the HTR process. This is a bottle-neck which is holding back the massive industrial document processing. Baseline detection can be used not only to segment page images into line images but also for many other document processing steps. Baseline detection problem can be formulated as a clustering problem over a set of interest points. In this work, we study the use of an automatic baseline detection technique, based on interest point clustering, in Arabic handwritten documents. The experiments reveal that this technique provides promising results for this task.

CCREA: Creación Colaborativa en Red de contenidos para Exámenes por el Alumnado

Tradicionalmente se ha cambiado el papel pasivo del alumno frente a la evaluacion involucrandole de dos maneras En las fases de correccion posterior mediante tecnicas de correccion interpares, correccion ciega, revision de examenes u otras. Las tutorias y las resoluciones de dudas previas al examen tambien mejoran el papel activo del alumno. CCREA es una metodologia de trabajo que involucra al alumno en el proceso colaborativo de elaboracion de los contenidos que potencialmente pudieran aparecer en las pruebas evaluadoras de las asignaturas en las que esta matriculado. CCREA se enmarca dentro de la Evaluacion Participativa empleando herramientas abiertas en red para crear tanto los contenidos de las pruebas objetivas como los mecanismos para la colaboracion y comunicacion entre todas las partes interesadas en el proceso creativo del examen.

Spontaneous handwriting text recognition and classification using finite-state models

Finite-state models are used to implement a handwritten text recognition and classification system for a real application entailing casual, spontaneous writing with large vocabulary. Handwritten short phrases which involve a wide variety of writing styles and contain many non-textual artifacts, are to be classified into a small number of predefined classes. To this end, two different types of statistical framework for phrase recognition-classification are considered, based on finite-state models. HMMs are used for text recognition process. Depending to the considered architecture, N-grams are used for performing text recognition and then text classification (serial approach) or for performing both simultaneously (integrated approach). The multinomial text classifier is also employed in the classification phase of the serial approach. Experimental results are reported which, given the extreme difficulty of the task, are encouraging.