Error handling approach using characterization and correction steps for handwritten document analysis

Quiniou, Solen; Cheriet, Mohamed; Anquetil, Éric

doi:10.1007/s10032-011-0156-6

Cited by 3 publications

(3 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To this purpose, active learning techniques can be used [26]. In our approach, we adopt a strategy based on the use of confidence measures [19,31] in order to select which words should be supervised [27].…”

Section: Confidence Measuresmentioning

confidence: 99%

Interactive handwriting recognition with limited user effort

Serrano

Giménez

Civera

et al. 2013

IJDAR

View full text Add to dashboard Cite

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.

show abstract

Section: Confidence Measuresmentioning

confidence: 99%

Interactive handwriting recognition with limited user effort

Serrano

Giménez

Civera

et al. 2013

IJDAR

View full text Add to dashboard Cite

show abstract

“…Does applying OCR postcorrection to the output of a multilingual HTR model improve its performance? At the time of the experiment, no HTR postcorrection software was publicly available, with research on this topic being scarce [15,16]. As implementing an HTR postcorrection algorithm from scratch is out of the scope of this study, OCR postcorrection was used instead.…”

mentioning

confidence: 99%

Experimenting with Training a Neural Network in Transkribus to Recognise Text in a Multilingual and Multi-Authored Manuscript Collection

Capurro,

Provatorova,

Kanoulas

2023

Heritage

View full text Add to dashboard Cite

This work aims at developing an optimal strategy to automatically transcribe a large quantity of uncategorised, digitised archival documents when resources include handwritten text by multiple authors and in several languages. We present a comparative study to establish the efficiency of a single multilingual handwritten text recognition (HTR) model trained on multiple handwriting styles instead of using a separate model for every language. When successful, this approach allows us to automate the transcription of the archive, reducing manual annotation efforts and facilitating information retrieval. To train the model, we used the material from the personal archive of the Dutch glass artist Sybren Valkema (1916–1996), processing it with Transkribus.

show abstract

“…For instance, Quiniou et al [2011] propose a technique to improve the performance of a HTR system by obtaining a consensus hypothesis out of a n-best lists, and then, characterizing the errors and correcting them. Similarly, Farooq et al [2009] use a translation model to conduct an automatically postediting.…”

Section: Introductionmentioning

confidence: 99%

Multimodal interactive structured prediction.

Gonzalvo¹

View full text Add to dashboard Cite

Document prepared and typeset in L A T E X. Cover design by Vicent Alabau. Original photo of the Hypertext Editing System console in use at Brown University by Greg Lloyd. Retrieved from Wikimedia Commons with license CC-BY-2.0.You are free to share (copy and redistribute the material in any medium or format) and adapt (remix, transform, and build upon the material for any purpose) under the following terms: You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. AcknowledgementsVolia aprofitar estes línies per a agrair a totes les persones que, d'una o altra forma, m'han oferit el seu suport i confiança per a que esta tesi s'haja pogut acabar.En primer lloc, voldria agrair sincerament a tots aquells que han sigut una referència a l'hora de formar-me com a investigador. En especial als meus directors de tesi, Paco i Alberto, i al que ho va ser en una etapa inicial, Carlos, pel seu suport al llarg dels anys. Ells varen confiar en mi i em donar l'oportunitat de entrar en este món tan apassionant. També voldria agrair a Alfons i a Enrique, amb els quals vaig fer el projecte fi de carrera, per introduir-me en l'àrea de reconeixement de formes i reclutar-me al grup PRHLT. Tots ells, junt als altres professors del PRHLT, m'han ensenyat la importància de la rigorositat científica, el desenvolupament teòric, i en resum, tot allò qué suposa la professionalitat al món científic. I also would like to thank Prof. Hermann Ney for welcoming me in his research group for three months. During that period I had the chance to learn how is organized one of the top research groups in pattern recognition. Thanks to David Vilar and Daniel Stein for making my stay in Aachen much more enjoyable. Of course, I also would like to thank Prof. José Oncina, Prof. Philippe Langlais and Prof. Philipp Koehn for their kind reviews and valuable comments regarding the draft of this thesis, and Prof. Enrique Vidal, Prof. Marcello Federico and Prof. Philipp Koehn (again) for being members of the board committee. Special thanks to José Oncina, whose work has been a great inspiration for many of the work carried out in this thesis.Al llarg d'estos anys com a informàtic i com investigador he anat trobant companys d'estudis i de feina que han acabat esdevenint amics. Tot va començar amb els Jailers, amb qui vaig descobrir que la informàtica podia ser una cosa molt divertida. Anys després, el laboratori L105 va resultar ser no sols un lloc on creixer professionalment sinó també un lloc on creixer personalment. Gràcies a Neus, Stella, Maite, Vero i Miriam per escoltar les meues penes, i per l'afecte que m'han donat. A German, Jorge, Guillem i la resta del laboratori per compartir les rises i els bugs del compilador. Més endavant, vaig acabar a l'ITI junt a alguns companys com Vero i German. Allí vaig obrir una nova etapa iii professional amb el projecte MIPRCV i ha continuat amb CasMaCat. Este periode ...

show abstract

Error handling approach using characterization and correction steps for handwritten document analysis

Cited by 3 publications

References 37 publications

Interactive handwriting recognition with limited user effort

Interactive handwriting recognition with limited user effort

Experimenting with Training a Neural Network in Transkribus to Recognise Text in a Multilingual and Multi-Authored Manuscript Collection

Multimodal interactive structured prediction.

Contact Info

Product

Resources

About