JoinPap V2: a Human-Centric AI Approach for Papyrus Fragment Reassembly

Reconstructing large-scale textual artifacts such as fragmented ancient papyri remains a major challenge in digital humanities, requiring labor-intensive expert intervention. This paper introduces a new interactive AI-based system, JoinPap V2 , that follows a Human-Centric AI (HCAI) approach to support experts in this complex task. We employ a deep binary classifier to evaluate the plausibility of a proposed join by simultaneously checking the geometric fit of fragment borders and the continuity of visual content (e.g., fiber structure and ink traces). A key component of the approach is a data preparation pipeline that generates virtual positive tears to mitigate data scarcity. We integrate this fragment matching model into the JoinPap interface, which aggregates scores across both the fragment recto and verso and visualizes plausible joins together with their confidence. We provide a quantitative assessment of the model on a validation set, showing its ability to distinguish positive and negative joins. We also report qualitative feedback from expert papyrologists on the interface, together with a case-study evaluation of AI-suggested joins on two challenging fragment rolls, showing promising improvements in the ranking of plausible join proposals and demonstrating the system's potential to support papyrological reassembly work.

Authors

Institutions

Publication Details

Journal
Journal on Computing and Cultural Heritage
Published
2026-11-16
DOI
https://doi.org/10.1145/3847117
Primary Topic
Image Processing and 3D Reconstruction
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

JoinPap V2: a Human-Centric AI Approach for Papyrus Fragment Reassembly

Massimiliano Corsini, Nicola Messina, Fabrizio Falchi, Paolo Cignoni et al.
Journal on Computing and Cultural Heritage
Image Processing and 3D Reconstruction
article

JoinPap V2: a Human-Centric AI Approach for Papyrus Fragment Reassembly

Massimiliano Corsini, Nicola Messina, Fabrizio Falchi, Paolo Cignoni, Francesca Maltomini, Fabio Carrara, Valentina Iannace, Giulia Mirante
article en

Abstract

Reconstructing large-scale textual artifacts such as fragmented ancient papyri remains a major challenge in digital humanities, requiring labor-intensive expert intervention. This paper introduces a new interactive AI-based system, JoinPap V2 , that follows a Human-Centric AI (HCAI) approach to support experts in this complex task. We employ a deep binary classifier to evaluate the plausibility of a proposed join by simultaneously checking the geometric fit of fragment borders and the continuity of visual content (e.g., fiber structure and ink traces). A key component of the approach is a data preparation pipeline that generates virtual positive tears to mitigate data scarcity. We integrate this fragment matching model into the JoinPap interface, which aggregates scores across both the fragment recto and verso and visualizes plausible joins together with their confidence. We provide a quantitative assessment of the model on a validation set, showing its ability to distinguish positive and negative joins. We also report qualitative feedback from expert papyrologists on the interface, together with a case-study evaluation of AI-suggested joins on two challenging fragment rolls, showing promising improvements in the ranking of plausible join proposals and demonstrating the system's potential to support papyrological reassembly work.

Journal on Computing and Cultural Heritage
Istituto di Scienza e Tecnologie dell'Informazione "Alessandro Faedo" (IT)
Image Processing and 3D Reconstruction
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon