@inproceedings{xia-etal-2026-lvlm,
title = "{LVLM} Optimization for {A}ncient {C}hinese Book Image Analysis with Task-specific Augmentation and Instruction Tuning",
author = "Xia, Tian and
Liu, Yulong and
Wang, Yilin and
Yang, Yumeng and
Cai, Dongheng and
Tan, Yuyang and
Yang, Menghui",
editor = "Sprugnoli, Rachele and
Passarotti, Marco",
booktitle = "Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages ({LT}4{HALA} 2026) @ {LREC} 2026",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.lt4hala-1.30/",
doi = "10.63317/3w5rwjqr49n7",
pages = "299--304",
abstract = "Ancient Chinese text digitization faces challenges like variant characters and complex layouts. Based on the EvaHan 2026 tasks, this study proposes an LVLM-based framework for printed/handwritten text recognition and layout analysis. To effectively adapt the Qwen2.5-VL-7B-Instruct model, our methodology innovates through a dual-level optimization strategy: distinct augmentation strategies are developed for OCR and layout tasks, while task-specific prompt templates are engineered to decouple text transcription from coordinate prediction. This combined approach significantly enhances overall task proficiency, achieving Character Error Rates of 0.0372 (printed) and 0.0823 (handwritten), alongside a mean average Precision of 0.2933 for layout analysis. Results show general LVLMs underperform in zero-shot ancient text tasks, but fine-tuning with tailored strategies significantly boosts performance and highlights their potential."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xia-etal-2026-lvlm">
<titleInfo>
<title>LVLM Optimization for Ancient Chinese Book Image Analysis with Task-specific Augmentation and Instruction Tuning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tian</namePart>
<namePart type="family">Xia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yulong</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yilin</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yumeng</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dongheng</namePart>
<namePart type="family">Cai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuyang</namePart>
<namePart type="family">Tan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Menghui</namePart>
<namePart type="family">Yang</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 Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026) @ LREC 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rachele</namePart>
<namePart type="family">Sprugnoli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marco</namePart>
<namePart type="family">Passarotti</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>Ancient Chinese text digitization faces challenges like variant characters and complex layouts. Based on the EvaHan 2026 tasks, this study proposes an LVLM-based framework for printed/handwritten text recognition and layout analysis. To effectively adapt the Qwen2.5-VL-7B-Instruct model, our methodology innovates through a dual-level optimization strategy: distinct augmentation strategies are developed for OCR and layout tasks, while task-specific prompt templates are engineered to decouple text transcription from coordinate prediction. This combined approach significantly enhances overall task proficiency, achieving Character Error Rates of 0.0372 (printed) and 0.0823 (handwritten), alongside a mean average Precision of 0.2933 for layout analysis. Results show general LVLMs underperform in zero-shot ancient text tasks, but fine-tuning with tailored strategies significantly boosts performance and highlights their potential.</abstract>
<identifier type="citekey">xia-etal-2026-lvlm</identifier>
<identifier type="doi">10.63317/3w5rwjqr49n7</identifier>
<location>
<url>https://aclanthology.org/2026.lt4hala-1.30/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>299</start>
<end>304</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T LVLM Optimization for Ancient Chinese Book Image Analysis with Task-specific Augmentation and Instruction Tuning
%A Xia, Tian
%A Liu, Yulong
%A Wang, Yilin
%A Yang, Yumeng
%A Cai, Dongheng
%A Tan, Yuyang
%A Yang, Menghui
%Y Sprugnoli, Rachele
%Y Passarotti, Marco
%S Proceedings of the Fourth Workshop on Language Technologies for Historical and Ancient Languages (LT4HALA 2026) @ LREC 2026
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F xia-etal-2026-lvlm
%X Ancient Chinese text digitization faces challenges like variant characters and complex layouts. Based on the EvaHan 2026 tasks, this study proposes an LVLM-based framework for printed/handwritten text recognition and layout analysis. To effectively adapt the Qwen2.5-VL-7B-Instruct model, our methodology innovates through a dual-level optimization strategy: distinct augmentation strategies are developed for OCR and layout tasks, while task-specific prompt templates are engineered to decouple text transcription from coordinate prediction. This combined approach significantly enhances overall task proficiency, achieving Character Error Rates of 0.0372 (printed) and 0.0823 (handwritten), alongside a mean average Precision of 0.2933 for layout analysis. Results show general LVLMs underperform in zero-shot ancient text tasks, but fine-tuning with tailored strategies significantly boosts performance and highlights their potential.
%R 10.63317/3w5rwjqr49n7
%U https://aclanthology.org/2026.lt4hala-1.30/
%U https://doi.org/10.63317/3w5rwjqr49n7
%P 299-304
Markdown (Informal)
[LVLM Optimization for Ancient Chinese Book Image Analysis with Task-specific Augmentation and Instruction Tuning](https://aclanthology.org/2026.lt4hala-1.30/) (Xia et al., LT4HALA 2026)
ACL