Call for proposal

Call for proposals to the conference "Ancient documents and automatic recognition of handwriting", 23 and 24 June 2022, ENC, Paris. 

This scientific event will be held on 23 and 24 June 2022 at the École nationale des chartes, Paris. This event will be held in a mixed format with a broadcast on Youtube. 

Today, many projects include an automatic text acquisition step in their production chain or data exploitation. Several transcription platforms and different HTR engines are now available. The integration of this technology into more and more efficient processing chains has led to the automation of tasks that question the place of researchers in the text production process. This new, data-intensive practice makes it urgent to gather, and therefore harmonise corpora for the constitution of training corpora, but also to make them available in order to improve the quality of HTR results. 

Thus, the École nationale des chartes, in partnership with the LabEx Hastec and the LAMOP, through CREMMALab project supported by the DIM MAP, will organise on 23 and 24 June 2022 two days of conferences combining philological and technical questions on the use of HTR for ancient documents. We will take this opportunity to review HTR through its tools, results and new practices induced by its use in publishing and exploitation projects. We hope that this event will also be an opportunity to bring together a growing international community of researchers to discuss the use of HTR in their scientific projects. 

The proposal is based on a desire to treat the theme of this conference from a technical point of view, while linking it to scientific problem of constitution and/or exploitation of corpor, by questioning both the practical aspects of the use of this technology (development of HTR engine, transcription interface, user interface to use and train models, etc.), while raising its methodological issues and its impact on research data. 

We propose three different axes for the talks : 

-Axis 1: Constitution of training data and HTR models (segmentation, transcription, constitution of corpora, sharing, etc.)

-Axis 2: Machine learning, technologies, transcription platforms (presentation of HTR tools and their performances, presentation of transcription assistance tools)

-Axis 3: Projects or editorial chains using HTR; exploitations of HTR data. 

We hope that the communications and exchanges during this conference will lead to a publication.

Proposals should be between 750 and 1000 words (excluding bibliography and figure). Proposals for contributions can be made in French or English until 18 February 2022.

Organising Committee: Ariane Pinche and Florianne Chiffoleau

Scientific Committee: Jean-Baptiste Camps, Alix Chagué, Thibault Clérice, Frédéric Duval, Vincent Jolivet, Benjamin Kiessling, Nicolas Perreaux, Ariane Pinche, Laurent Romary, Peter Stokes

 For any information request, you can contact us at the following address: dahtr@sciencesconf.org