Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 12
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Arabic Printed Word Recognition Using Windowed Bernoulli HMMs

  • Alkhoury, Ihab
  • Andrés Ferrer, Jesús
  • Giménez Pastor, Adrián
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Hidden Markov Models (HMMs) are now widely used for off-line text recognition in many languages and, in particular, Arabic. In previous work, we proposed to directly use columns of raw, binary image pixels, which are directly fed into embedded Bernoulli (mixture) HMMs, that is, embedded HMMs in which the emission probabilities are modeled with Bernoulli mixtures. The idea was to by-pass feature extraction and to ensure that no discriminative information is filtered out during feature extraction, which in some sense is integrated into the recognition model. More recently, we extended the column bit vectors by means of a sliding window of adequate width to better capture image context at each horizontal position of the word image. However, these models might have limited capability to properly model vertical image distortions. In this paper, we have considered three methods of window repositioning after window extraction to overcome this limitation. Each sliding window is translated (repositioned) to align its center to the center of mass. Using this approach, state-of-art results are reported on the Arabic Printed Text Recognition (APTI) database., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no 287755. Also supported by the Spanish Government (Plan E, iTrans2 TIN2009-14511 and AECID 2011/2012 grant).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A System Architecture to Support Cost-Effective Transcription and Translation of Large Video Lecture Repositories

  • Silvestre Cerdà, Joan Albert|||0000-0003-2291-8296
  • Jiménez López, Manuel
  • Turró Ribalta, Carlos|||0000-0002-3840-9405
  • Juan, Alfons|||0000-0002-9984-4072
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Pérez González de Martos, Alejandro Manuel
[EN] Online video lecture repositories are rapidly growing and becoming established as fundamental knowledge assets. However, most lectures are neither transcribed nor translated because of the lack of cost-effective solutions that can give accurate enough results. In this paper, we describe a system architecture that supports the cost-effective transcription and translation of large video lecture repositories. This architecture has been adopted in the EU project transLectures and is now being tested on a repository of more than 9000 video lectures at the Universitat Politecnica de Valencia. Following a brief description of this repository and of the transLectures project, we describe the proposed system architecture in detail. We also report empirical results on the quality of the transcriptions and translations currently being maintained and steadily improved., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 287755. Funding was also provided by the Spanish Government with the FPU scholarship AP2010-4349.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Window repositioning for Printed Arabic Recognition

  • Alkhoury, Ihab
  • Andrés Ferrer, Jesús
  • Giménez Pastor, Adrián
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Bernoulli HMMs are conventional HMMs in which the emission probabilities are modeled with Bernoulli mixtures. They have recently been applied, with good results, in off-line text recognition in many languages, in particular, Arabic. A key idea that has proven to be very effective in this application of Bernoulli HMMs is the use of a sliding window of adequate width for feature extraction. This idea has allowed us to obtain very competitive results in the recognition of both Arabic handwriting and printed text. Indeed, a system based on it ranked first at the ICDAR 2011 Arabic recognition competition on the Arabic Printed Text Image (APTI) database. More recently, this idea has been refined by using repositioning techniques for extracted windows, leading to further improvements in Arabic handwriting recognition. In the case of printed text, this refinement led to an improved system which ranked second at the ICDAR 2013 second competition on APTI, only at a marginal distance from the best system. In this work, we describe the development of this improved system. Following evaluation protocols similar to those of the competitions on APTI, exhaustive experiments are detailed from which state-of-the-art results are obtained., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/ICT-287755) under grant agreement no. 287755. The research is also supported by the Spanish Government (Plan E, iTrans2 TIN2009-14511 and AECID 2011/2012 grant).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Integrating a State-of-the-Art ASR System into the Opencast Matterhorn Platform

  • Valor Miró, Juan Daniel
  • Pérez González de Martos, Alejandro Manuel
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Juan, Alfons|||0000-0002-9984-4072
[EN] In this paper we present the integration of a state-of-the-art ASR system into the Opencast Matterhorn platform, a free, open-source platform to support the management of educational audio and video content. The ASR system was trained on a novel large speech corpus, known as poliMedia, that was manually transcribed for the European project transLectures. This novel corpus contains more than 115 hours of transcribed speech that will be available for the research community. Initial results on the poliMedia corpus are also reported to compare the performance of different ASR systems based on the linear interpolation of language models. To this purpose, the in-domain poliMedia corpus was linearly interpolated with an external large-vocabulary dataset, the well-known Google N-Gram corpus. WER figures reported denote the notable improvement over the baseline performance as a result of incorporating the vast amount of data represented by the Google N-Gram corpus., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no 287755. Also supported by the Spanish Government (MIPRCV ”Consolider Ingenio 2010” and iTrans2 TIN2009-14511) and the Generalitat Valenciana (Prometeo/2009/014).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Efficiency and usability study of innovative computer-aided transcription strategies for video lecture repositories

  • Valor Miró, Juan Daniel
  • Silvestre Cerdà, Joan Albert|||0000-0003-2291-8296
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Turró Ribalta, Carlos|||0000-0002-3840-9405
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Video lectures are widely used in education to support and complement face-to-face lectures. However, the utility of these audiovisual assets could be further improved by adding subtitles that can be exploited to incorporate added-value functionalities such as searchability, accessibility, translatability, note-taking, and discovery of content-related videos, among others. Today, automatic subtitles are prone to error, and need to be reviewed and post-edited in order to ensure that what students see on-screen are of an acceptable quality. This work investigates different user interface design strategies for this post-editing task to discover the best way to incorporate automatic transcription technologies into large educational video repositories. Our three-phase study involved lecturers from the Universitat Polite`cnica de Vale`ncia (UPV) with videos available on the poliMedia video lecture repository, which is currently over 10,000 video objects. Simply by conventional post-editing automatic transcriptions users almost reduced to half the time that would require to generate the transcription from scratch. As expected, this study revealed that the time spent by lecturers reviewing automatic transcriptions correlated directly with the accuracy of said transcriptions. However, it is also shown that the average time required to perform each individual editing operation could be precisely derived and could be applied in the definition of a user model. In addition, the second phase of this study presents a transcription review strategy based on confidence measures (CM) and compares it to the conventional post-editing strategy. Finally, a third strategy resulting from the combination of that based on CM with massive adaptation techniques for automatic speech recognition (ASR), achieved to improve the transcription review efficiency in comparison with the two aforementioned strategies. 2015 Elsevier B.V. All rights reserved., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under Grant agreement no. 287755 (transLectures) and ICT Policy Support Programme (ICT PSP/2007-2013) as part of the Competitiveness and Innovation Framework Programme (CIP) under Grant agreement no. 621030 (EMMA), and the Spanish MINECO Active2Trans (TIN2012-31723) research project.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Explicit length modelling for statistical machine translation

  • Silvestre Cerdà, Joan Albert|||0000-0003-2291-8296
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Andrés Ferrer, Jesús
[EN] Explicit length modelling has been previously explored in statistical pattern recognition with successful results. In this paper, two length models along with two parameter estimation methods and two alternative parametrisations for statistical machine translation (SMT) are presented. More precisely, we incorporate explicit bilingual length modelling in a state-of-the-art log-linear SMT system as an additional feature function in order to prove the contribution of length information. Finally, a systematic evaluation on reference SMT tasks considering different language pairs proves the benefits of explicit length modelling., Work supported by the EC (FEDER/FSE) under the transLectures project (FP7-ICT-2011-7-287755) and the Spanish MEC/MICINN under the MIPRCV "Consolider Ingenio 2010" program (CSD2007-00018) and iTrans2 (TIN2009-14511) projects and FPU grant (AP2010-4349). Also supported by the Spanish MITyC under the erudito.com (TSI-020110-2009-439) project and by the Generalitat Valenciana under grants Prometeo/2009/014 and GV/2010/067, and by the UPV under the AdInTAO (20091027) project. The authors wish to thank the anonymous reviewers for their criticisms and suggestions.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Speaker-adapted confidence measures for speech recognition of video lectures

  • Sanchez-Cortina, Isaias
  • Andrés Ferrer, Jesús
  • Sanchis Navarro, José Alberto|||0000-0002-2943-0990
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Automatic speech recognition applications can benefit from a confidence measure (CM) to predict the reliability of the output. Previous works showed that a word-dependent native Bayes (NB) classifier outperforms the conventional word posterior probability as a CM. However, a discriminative formulation usually renders improved performance due to the available training techniques. Taking this into account, we propose a logistic regression (LR) classifier defined with simple input functions to approximate to the NB behaviour. Additionally, as a main contribution, we propose to adapt the CM to the speaker in cases in which it is possible to identify the speakers, such as online lecture repositories. The experiments have shown that speaker-adapted models outperform their non-adapted counterparts on two difficult tasks from English (videoLectures.net) and Spanish (poliMedia) educational lectures. They have also shown that the NB model is clearly superseded by the proposed LR classifier., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement no 287755. Also supported by the Spanish MINECO (iTrans2 TIN2009-14511 and Active2Trans TIN2012-31723) research projects and the FPI Scholarship BES-2010-033005.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Effective balancing error and user effort in interactive handwriting recognition

  • Serrano Martinez Santos, Nicolas
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Sanchis Navarro, José Alberto|||0000-0002-2943-0990
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Transcription of handwritten text documents is an expensive and time-consuming task. Unfortunately, the accuracy of current state-of-the-art handwriting recognition systems cannot guarantee fully-automatic high quality transcriptions, so we need to revert to the computer assisted approach. Although this approach reduces the user effort needed to transcribe a given document, the transcription of handwriting text documents still requires complete manual supervision. An especially appealing scenario is the interactive transcription of handwriting documents, in which the user defines the amount of errors that can be tolerated in the final transcribed document. Under this scenario, the transcription of a handwriting text document could be obtained efficiently, supervising only a certain number of incorrectly recognised words. In this work, we develop a new method for predicting the error rate in a block of automatically recognised words, and estimate how much effort is required to correct a transcription to a certain user-defined error rate. The proposed method is included in an interactive approach to transcribing handwritten text documents, which efficiently employs user interactions by means of active and semi-supervised learning techniques, along with a hypothesis recomputation algorithm based on constrained Viterbi search. Transcription results, in terms of trade-off between user effort and transcription accuracy, are reported for two real handwritten documents, and prove the effectiveness of the proposed approach., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under Grant agreement No 287755 (transLectures). Also supported by the EC (FEDER, FSE), the Spanish Government (MICINN, MITyC, "Plan E", under grants MIPRCV "Consolider Ingenio 2010", MITTRAL (TIN2009-14633-C03-01), iTrans2 (TIN2009-14511), and FPU (AP2007-02867), and the Generalitat Valenciana (Grants Prometeo/2009/014 and GV/2010/067). Special thanks to Jesus Andres for his fruitful discussions.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Interactive handwriting recognition with limited user effort

  • Serrano Martinez Santos, Nicolas
  • Giménez Pastor, Adrián
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Sanchis Navarro, José Alberto|||0000-0002-2943-0990
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Transcription of handwritten text in (old) documents is an important, time-consuming task for digital libraries. Although post-editing automatic recognition of handwritten text is feasible, it is not clearly better than simply ignoring it and transcribing the document from scratch. A more effective approach is to follow an interactive approach in which both the system is guided by the user, and the user is assisted by the system to complete the transcription task as efficiently as possible. Nevertheless, in some applications, the user effort available to transcribe documents is limited and fully supervision of the system output is not realistic. To circumvent these problems, we propose a novel interactive approach which efficiently employs user effort to transcribe a document by improving three different aspects. Firstly, the system employs a limited amount of effort to solely supervise recognised words that are likely to be incorrect. Thus, user effort is efficiently focused on the supervision of words for which the system is not confident enough. Secondly, it refines the initial transcription provided to the user by recomputing it constrained to user supervisions. In this way, incorrect words in unsupervised parts can be automatically amended without user supervision. Finally, it improves the underlying system models by retraining the system from partially supervised transcriptions. In order to prove these statements, empirical results are presented on two real databases showing that the proposed approach can notably reduce user effort in the transcription of handwritten text in (old) documents., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under Grant Agreement No 287755 (transLectures). Also supported by the Spanish Government (MICINN, MITyC, "Plan E", under Grants MIPRCV "Consolider Ingenio 2010", MITTRAL (TIN2009-14633-C03-01), erudito.com (TSI-020110-2009-439), iTrans2 (TIN2009-14511), and FPU (AP2007-02867), and the Generalitat Valenciana (Grants Prometeo/2009/014 and GV/2010/067).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Speaker-Adapted Confidence Measures for ASR using Deep Bidirectional Recurrent Neural Networks

  • Del Agua Teba, Miguel Angel
  • Giménez Pastor, Adrián
  • Sanchis Navarro, José Alberto|||0000-0002-2943-0990
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Juan, Alfons|||0000-0002-9984-4072
[EN] In the last years, Deep Bidirectional Recurrent Neural Networks (DBRNN) and DBRNN with Long Short-Term Memory cells (DBLSTM) have outperformed the most accurate classifiers for confidence estimation in automatic speech recognition. At the same time, we have recently shown that speaker adaptation of confidence measures using DBLSTM yields significant improvements over non-adapted confidence measures. In accordance with these two recent contributions to the state of the art in confidence estimation, this paper presents a comprehensive study of speaker-adapted confidence measures using DBRNN and DBLSTM models. Firstly, we present new empirical evidences of the superiority of RNN-based confidence classifiers evaluated over a large speech corpus consisting of the English LibriSpeech and the Spanish poliMedia tasks. Secondly, we show new results on speaker-adapted confidence measures considering a multi-task framework in which RNN-based confidence classifiers trained with LibriSpeech are adapted to speakers of the TED-LIUM corpus. These experiments confirm that speaker-adapted confidence measures outperform their non-adapted counterparts. Lastly, we describe an unsupervised adaptation method of the acoustic DBLSTM model based on confidence measures which results in better automatic speech recognition performance., This work was supported in part by the European Union's Horizon 2020 research and innovation programme under Grant 761758 (X5gon), in part by the Seventh Framework Programme (FP7/2007-2013) under Grant 287755 (transLectures), in part by the ICT Policy Support Programme (ICT PSP/2007-2013) as part of the Competitiveness and Innovation Framework Programme under Grant 621030 (EMMA), and in part by the Spanish Government's TIN2015-68326-R (MINECO/FEDER) research project MORE.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Character-Based Handwritten Text Recognition of Multilingual Documents

  • Del Agua Teba, Miguel Angel
  • Serrano Martinez Santos, Nicolas
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Juan, Alfons|||0000-0002-9984-4072
[EN] An effective approach to transcribe handwritten text documents is to follow a sequential interactive approach. During the supervision phase, user corrections are incorporated into the system through an ongoing retraining process. In the case of multilingual documents with a high percentage of out-of-vocabulary (OOV) words, two principal issues arise. On the one hand, a minor yet important matter for this interactive approach is to identify the language of the current text line image to be transcribed, as a language dependent recognisers typically performs better than a monolingual recogniser. On the other hand, word-based language models suffer from data scarcity in the presence of a large number of OOV words, degrading their estimation and affecting the performance of the transcription system. In this paper, we successfully tackle both issues deploying character-based language models combined with language identification techniques on an entire 764-page multilingual document. The results obtained significantly reduce previously reported results in terms of transcription error on the same task, but showed that a language dependent approach is not effective on top of character-based recognition of similar languages., The research leading to these results has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement n◦ 287755. Also supported by the Spanish Government (MIPRCV ”Consolider Ingenio 2010”, iTrans2 TIN2009-14511, MITTRAL TIN2009-14633-C03-01 and FPU AP2007-0286) and the Generalitat Valenciana (Prometeo/2009/014).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Multilingual videos for MOOCs and OER

  • Valor Miró, Juan Daniel
  • Baquero-Arnal, Pau
  • Civera Saiz, Jorge|||0000-0002-0963-0143
  • Turró Ribalta, Carlos|||0000-0002-3840-9405
  • Juan, Alfons|||0000-0002-9984-4072
[EN] Massive Open Online Courses (MOOCs) and Open Educational Resources (OER) are rapidly growing, but are not usually offered in multiple languages due to the lack of cost-effective solutions to translate the different objects comprising them and particularly videos. However, current state-of-the-art automatic speech recognition (ASR) and machine translation (MT) techniques have reached a level of maturity which opens the possibility of producing multilingual video subtitles of publishable quality at low cost. This work summarizes authors' experience in exploring this possibility in two real-life case studies: a MOOC platform and a large video lecture repository. Apart from describing the systems, tools and integration components employed for such purpose, a comprehensive evaluation of the results achieved is provided in terms of quality and efficiency. More precisely, it is shown that draft multilingual subtitles produced by domainadapted ASR/MT systems reach a level of accuracy that make them worth post-editing, instead of generating them ex novo, saving approximately 25%-75% of the time. Finally, the results reported on user multilingual data consumption reflect that multilingual subtitles have had a very positive impact in our case studies boosting student enrolment, in the case of the MOOC platform, by 70% relative., The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 287755 (transLectures) and from the EU's ICT Policy Support Programme as part of the Competitiveness and Innovation Framework Programme under grant agreement no. 621030 (EMMA). Additionally, it is supported by the Spanish research project TIN2015-68326-R (MINECO/FEDER).