Revisiting N-Gram Models: Their Impact in Modern Neural Networks for Handwritten Text Recognition

Autor: Tarride, Solène, Kermorvant, Christopher

In recent advances in automatic text recognition (ATR), deep neural networks have demonstrated the ability to implicitly capture language statistics, potentially reducing the need for traditional language models. This study directly addresses whether

Externí odkaz: http://arxiv.org/abs/2404.19317

The Socface Project: Large-Scale Collection, Processing, and Analysis of a Century of French Censuses

Autor: Boillet, Mélodie, Tarride, Solène, Blanco, Manon, Rigal, Valentin, Schneider, Yoann, Abadie, Bastien, Kesztenbaum, Lionel, Kermorvant, Christopher

This paper presents a complete processing workflow for extracting information from French census lists from 1836 to 1936. These lists contain information about individuals living in France and their households. We aim at extracting all the informatio

Externí odkaz: http://arxiv.org/abs/2404.18706

Handwritten Text Recognition from Crowdsourced Annotations

Autor: Tarride, Solène, Faine, Tristan, Boillet, Mélodie, Mouchère, Harold, Kermorvant, Christopher

In this paper, we explore different ways of training a model for handwritten text recognition when multiple imperfect or noisy transcriptions are available. We consider various training configurations, such as selecting a single transcription, retain

Externí odkaz: http://arxiv.org/abs/2306.10878

SIMARA: a database for key-value information extraction from full pages

Autor: Tarride, Solène, Boillet, Mélodie, Moufflet, Jean-François, Kermorvant, Christopher

We propose a new database for information extraction from historical handwritten documents. The corpus includes 5,393 finding aids from six different series, dating from the 18th-20th centuries. Finding aids are handwritten documents that contain met

Externí odkaz: http://arxiv.org/abs/2304.13606

A comparative study of information extraction strategies using an attention-based neural network

Autor: Tarride, Solène, Lemaitre, Aurélie, Coüasnon, Bertrand, Tardivel, Sophie

Publikováno v: 15th IAPR International Workshop on Document Analysis Systems
15th IAPR International Workshop on Document Analysis Systems, May 2022, La Rochelle, France
Document Analysis Systems ISBN: 9783031065545

International audience; This article focuses on information extraction in historical handwritten marriage records. Traditional approaches rely on a sequential pipeline of two consecutive tasks: handwriting recognition is applied before named entity r

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::9fc4db61ebfea392fc7d58e5c55547b9
https://hal.science/hal-03677908