Research scientist at DHLAB @EPFL on @ImpressoProject | Information Extraction | Citation Mining | Digital Classics | Formerly @dai_weltweit and @KingsDH

Joined June 2009
132 Photos and videos
Matteo Romanello retweeted
We have three Swiss-based openings at Odoma, we're very excited to grow the company and look forward to have new people joining us: odoma.ch/openjobs. We are looking for a software developer, a machine learning engineer, and a project manager. We are very flexible re. seniority, salary, and remote working. Please get in touch for questions, and consider applying or reposting. @mr56k @laurent_bolli.
7
7
1,190
Want to hear about large dataset and generic models for handwritten text recognition of medieval manuscripts ? We talk about it at #Dh2024 at 8.30 Want to hear about standardisation efforts for layout recognition with TEI building as a downstream task ? We talk about it at 10.30
1
5
17
1,137
"We believe we can represent the vast majority of services from Perseus 4 in such a minimal computing framework. Such a version of Perseus could be downloaded and run locally." I agree that there is a real space/need/opportunity for minimal computing approaches in DH projects.
Problems with Perseus 4 and work towards Perseus 6 » Perseus Digital Library Updates sites.tufts.edu/perseusupdat… via @PerseusDigLib
187
Matteo Romanello retweeted
Introducing Florence 2 for Medieval Zone Detection, now available on @huggingface ! This model was trained on the CATMuS Medieval Segmentation dataset, also available on HuggingFace. It uses the SegmOnto format and trained on 12 zones (some with very few examples). I had a lot of fun working on this model and learned quite a bit. I'll be speaking about this in more depth in a forthcoming blog. #Medieval #DataScience #MachineLearning
1
11
69
4,739
Two documents already submitted. If you'd like documents in Latin, specifically documents of practice, transcribed in the CATMuS dataset, sent them our way ! x.com/PonteIneptique/status/…

HTRogene proj. is looking for ℹ️ docs of practice (pref. cursive) 🇻🇦 in Latin 🏰 date 1000-1550 🆓 open license for the images 🪪 with metadata (shelfmark, century). So that we transcribe 5/10 pages. Anyone wants something transcribed ? :) forms.gle/4MxLQg97eQwi54xX6 <3 RT
2
2
300
Matteo Romanello retweeted
#bibliodata business ready to be introduced at this year's @DARIAHeu annual event. Come to discuss and share the experiences on @SSHOpenCloud Open Marketplace workflows within the dedicated panel today at 2 PM in Auditorium B2! #DARIAH2024
11
23
1,244
Matteo Romanello retweeted
Help us spread the word: #CfP for the first Impresso conference! "#Transmedia History: Circulations, Reconfigurations and New Methodologies" on 27 and 28 January 2025 in Lausanne @unil - the submission deadline is 15.7. and all infos are here: impresso.github.io/transmedi…
17
20
1,635
Matteo Romanello retweeted
#johdpapers New #datapaper out 📢 "Translation Alignment for #AncientGreek: Annotation Guidelines and Gold Standards" ✍️@geophilologist @farnooshamsian @TariqYousef1 @rmavirumquecano Anise d’Orange Ferreira & Michel Ferreira dos Reis 📰👉rb.gy/sggl9d #CompLing
1
6
18
2,009
Matteo Romanello retweeted
19 Dec 2023
Call for Papers: Images Deluge & Globalization | June 19-21 2024 | Geneva, CH | submit your abstract before Feb 1st 2024. unige.ch/visualcontagions/te…
1
2
152
Matteo Romanello retweeted
#prixscienceouverte #donneesrecherche Catégorie Prix jeunes chercheurs HTR-United : catalogue pour partager et découvrir des données d'entraînement pour la transcription et la segmentation automatique de documents @Alix_Tz @PonteIneptique @Inria @sup_recherche #scienceouverte
1
15
20
3,961
Matteo Romanello retweeted
Replying to @CompHumResearch
@CompHumResearch 2023 proceedings are out! The scale of the conference exploded: 49 full papers across 830 pages 👾 ceur-ws.org/Vol-3558/ Not possible w/o @FolgertK, Melvin Wevers, @Iza_Romanowska @fotisja @F_Cafiero @MariePuren and all authors following CEUR guildelines 👀

21
29
3,774
Matteo Romanello retweeted
You know the holidays are coming, when Linked Pasts put up their wonderful programme! Check it out: activities from Dec 4-15th! linkedpasts.hcommons.org/202…
3
3
717
Matteo Romanello retweeted
Updates from @jacobwegner and github.com/pletcher in our ability to coordinate CTS textual data with annotations onto IIIF: beyond-translation.perseus.o…. Still a work in progress and you will need to fiddle a bit still but you now see different scholia and lemma in the Venetus A.
5
17
1,130
Matteo Romanello retweeted
Digital Classicist Seminar Wir legen noch einen drauf! Am 14.11. spricht Youssef Nader von der @FU_Berlin über sein prämiertes Verfahren zur Entzifferung verkohlter Papyri: bit.ly/469OsiK. @bbaw_de
4
9
2,641
Matteo Romanello retweeted
Big news: after years of organizing workshops, archaeologists and historians are joining forces to offer a summer school in network analysis! pastnetworks.net/ 25-29.06.2024, Barcelona with @martenduering @anduviera @tombrughmans @cprog7 @MattMazzamurro @HNR_org and myself
37
83
7,259
Matteo Romanello retweeted
Join us for the first DH Alumni event. Alumni will share with us their DH journey. The presentations will be followed by an "ask me anything" session, where our DH alumni will answer questions from the audience. The event will be rounded up with drinks and networking for all.
2
8
1,232
Matteo Romanello retweeted
Really nice @gradio appliation for doing Hand Written Text Recognition (#HTR/#OCR) shared by @ra_sverige. The app has a great overview of the steps involved in the pipeline and the demo is underpinned by openly shared models and datasets! huggingface.co/spaces/Riksar…
1
4
14
1,854
Matteo Romanello retweeted
Check our forthcoming workshop on AI and large language models to analyze literature and fictions. Attendance is free but registration mandatory. workshop-llms4cls.github.io #DH #AI #DHAI

1
17
40
14,917