@inproceedings{barmandah-etal-2026-ketaba,
title = "Ketaba-{OCR} at {AR}-{MS} {N}akba{NLP} 2026: Efficient Adaptation of Vision-Language Models for Handwritten Recognition",
author = "Barmandah, Hassan and
Eldin, Fatimah Emad and
Al Jallad, Khloud and
Nacar, Omer",
editor = "Jarrar, Mustafa and
El-Haj, Mo and
Haddad, Amal and
Atiani, Serin and
Abudalfa, Shadi and
Regier, Terry and
Rayson, Paul and
Sima{'}an, Khalil and
Mansour, Camille",
booktitle = "Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ {LREC} 2026",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.nakbanlp-1.21/",
doi = "10.63317/3inc2znes52o",
pages = "160--170",
abstract = "This paper presents Ketaba-OCR-LoRA, a system developed for the NakbaNLP 2026 Shared Task on Arabic Manuscript Understanding (Subtask 2), which targets the transcription of the historically significant Omar Al-Saleh Memoir Collection written in Ruq{'}ah and Naskh scripts. We propose a parameter-efficient adaptation of a publicly available pretrained Arabic-English Handwritten Text Recognition (HRT) model, originally trained on handwritten corpora including the Muharaf dataset. Instead of adapting general Vision-Language Models from scratch, we fine-tune the HRT backbone using Low-Rank Adaptation (LoRA) and 4-bit quantization (QLoRA), reducing memory requirements from 40GB to approximately 8GB. Our final submission combines multiple model variants through a novel Linear+Boost weighted ensemble strategy. Our approach achieves a CER of 0.0819 and WER of 0.2588 on the blind test set (per-line evaluation), ranking 1st on per-line evaluation; on the official corpus-wide leaderboard, we rank 3rd (CER 0.0938, WER 0.2996). This work demonstrates that specialized pretrained HRT models substantially outperform general-purpose Vision-Language Models for Arabic manuscript transcription, and that parameter-efficient fine-tuning provides a practical and reproducible approach for low-resource cultural heritage digitization."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="barmandah-etal-2026-ketaba">
<titleInfo>
<title>Ketaba-OCR at AR-MS NakbaNLP 2026: Efficient Adaptation of Vision-Language Models for Handwritten Recognition</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hassan</namePart>
<namePart type="family">Barmandah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fatimah</namePart>
<namePart type="given">Emad</namePart>
<namePart type="family">Eldin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Khloud</namePart>
<namePart type="family">Al Jallad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Omer</namePart>
<namePart type="family">Nacar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mustafa</namePart>
<namePart type="family">Jarrar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mo</namePart>
<namePart type="family">El-Haj</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Amal</namePart>
<namePart type="family">Haddad</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Serin</namePart>
<namePart type="family">Atiani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shadi</namePart>
<namePart type="family">Abudalfa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Terry</namePart>
<namePart type="family">Regier</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Paul</namePart>
<namePart type="family">Rayson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Khalil</namePart>
<namePart type="family">Sima’an</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Camille</namePart>
<namePart type="family">Mansour</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This paper presents Ketaba-OCR-LoRA, a system developed for the NakbaNLP 2026 Shared Task on Arabic Manuscript Understanding (Subtask 2), which targets the transcription of the historically significant Omar Al-Saleh Memoir Collection written in Ruq’ah and Naskh scripts. We propose a parameter-efficient adaptation of a publicly available pretrained Arabic-English Handwritten Text Recognition (HRT) model, originally trained on handwritten corpora including the Muharaf dataset. Instead of adapting general Vision-Language Models from scratch, we fine-tune the HRT backbone using Low-Rank Adaptation (LoRA) and 4-bit quantization (QLoRA), reducing memory requirements from 40GB to approximately 8GB. Our final submission combines multiple model variants through a novel Linear+Boost weighted ensemble strategy. Our approach achieves a CER of 0.0819 and WER of 0.2588 on the blind test set (per-line evaluation), ranking 1st on per-line evaluation; on the official corpus-wide leaderboard, we rank 3rd (CER 0.0938, WER 0.2996). This work demonstrates that specialized pretrained HRT models substantially outperform general-purpose Vision-Language Models for Arabic manuscript transcription, and that parameter-efficient fine-tuning provides a practical and reproducible approach for low-resource cultural heritage digitization.</abstract>
<identifier type="citekey">barmandah-etal-2026-ketaba</identifier>
<identifier type="doi">10.63317/3inc2znes52o</identifier>
<location>
<url>https://aclanthology.org/2026.nakbanlp-1.21/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>160</start>
<end>170</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Ketaba-OCR at AR-MS NakbaNLP 2026: Efficient Adaptation of Vision-Language Models for Handwritten Recognition
%A Barmandah, Hassan
%A Eldin, Fatimah Emad
%A Al Jallad, Khloud
%A Nacar, Omer
%Y Jarrar, Mustafa
%Y El-Haj, Mo
%Y Haddad, Amal
%Y Atiani, Serin
%Y Abudalfa, Shadi
%Y Regier, Terry
%Y Rayson, Paul
%Y Sima’an, Khalil
%Y Mansour, Camille
%S Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F barmandah-etal-2026-ketaba
%X This paper presents Ketaba-OCR-LoRA, a system developed for the NakbaNLP 2026 Shared Task on Arabic Manuscript Understanding (Subtask 2), which targets the transcription of the historically significant Omar Al-Saleh Memoir Collection written in Ruq’ah and Naskh scripts. We propose a parameter-efficient adaptation of a publicly available pretrained Arabic-English Handwritten Text Recognition (HRT) model, originally trained on handwritten corpora including the Muharaf dataset. Instead of adapting general Vision-Language Models from scratch, we fine-tune the HRT backbone using Low-Rank Adaptation (LoRA) and 4-bit quantization (QLoRA), reducing memory requirements from 40GB to approximately 8GB. Our final submission combines multiple model variants through a novel Linear+Boost weighted ensemble strategy. Our approach achieves a CER of 0.0819 and WER of 0.2588 on the blind test set (per-line evaluation), ranking 1st on per-line evaluation; on the official corpus-wide leaderboard, we rank 3rd (CER 0.0938, WER 0.2996). This work demonstrates that specialized pretrained HRT models substantially outperform general-purpose Vision-Language Models for Arabic manuscript transcription, and that parameter-efficient fine-tuning provides a practical and reproducible approach for low-resource cultural heritage digitization.
%R 10.63317/3inc2znes52o
%U https://aclanthology.org/2026.nakbanlp-1.21/
%U https://doi.org/10.63317/3inc2znes52o
%P 160-170
Markdown (Informal)
[Ketaba-OCR at AR-MS NakbaNLP 2026: Efficient Adaptation of Vision-Language Models for Handwritten Recognition](https://aclanthology.org/2026.nakbanlp-1.21/) (Barmandah et al., NakbaNLP 2026)
ACL