Org. EIDA team
Place: Observatoire de Paris, Salle du Levant
Date: Dec. 13, 2024
With the support of the AISSAI center, in the context of the Thematic semester Digital Humanities and Artificial Intelligence.
Historical documents, particularly those in the history of science, encompass a wide range of non-discursive elements, including tables, diagrams, maps, and various forms of illustrations. These non-discursive textual forms have their own distinct histories and present unique interpretive challenges for historians, especially in relation to the discursive portions of the documents. Often, one key element in understanding these relationships, rely in the text regions which are directly included as part and constituent of the tables, diagrams, maps and illustrations. These text regions, however, pose specific challenges for computer vision technologies and historical document analysis. They can be difficult to identify within documents, exhibit diverse reading orientations, and are often challenging to transcribe due to their brevity and reliance on ad-hoc or localized symbols. This international workshop brings together teams that are currently engaged in active research on these topics, with the aim of evaluating our findings and exploring potential collaborative efforts to address shared challenges.
Participants
Jade Norindr, Matthieu Husson, Florence Somer, Scott Trigg, Mathieu Aubry, Raphael Baena, Katherine McDonough, Alix Chagué,…
Program
9h30-10h45
Alix Chagué (Montreal, INRIA), ‘Transcription of discursive and non-discursive documents: connecting project specificities and generic frameworks’
11h-12h15
Katherine McDonough (Lancaster University), “Text on Maps with MapReader: Current Work and New Directions”
14h-15h15
Jade Norindr (CNRS, PSL-Paris Observatory), Raphael Baena (LIGM, ENPC), ‘Extracting text regions and reading direction in a global corpus of astronomical diagrams”
15h30-16h30
General discussion