@inproceedings{hudi-etal-2026-lectrans,
title = "{L}ec{T}rans: A Multimodal Translation Benchmark for Academic Lectures",
author = "Hudi, Frederikus and
Nanakorn, Nina and
Vasselli, Justin and
Sakai, Yusuke and
Kamigaito, Hidetaka and
Tamura, Akihiro and
Watanabe, Taro",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {IJCNLP}-{AACL} 2026",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-aacl.93/",
pages = "1872--1897",
ISBN = "979-8-89176-444-6",
abstract = "Existing machine translation and multimodal translation benchmarks mainly focus on text-only inputs or short examples with tightly aligned image-text pairs. Academic lectures present a different challenge: spoken explanations, presentation slides, and domain-specific terminology interact across long-form discourse. LecTrans is constructed from 383 hours of expert-taught online lectures and provides transcripts, presentation slides, and an eight-language parallel evaluation subset covering English and seven target languages. The 162-minute translation subset spans 17 subjects and contains human-corrected English transcripts, verified slide text, and human-verified translations of both transcripts and slides. We define lecture translation as two complementary tasks: transcript translation for spoken lecture content and slide translation for visually presented text. We report text-only and multimodal baselines, showing that noisy transcripts, loosely aligned slide context, and domain-specific content remain challenging for current systems. LecTrans provides a testbed for evaluating multimodal translation in long-form academic lectures."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hudi-etal-2026-lectrans">
<titleInfo>
<title>LecTrans: A Multimodal Translation Benchmark for Academic Lectures</title>
</titleInfo>
<name type="personal">
<namePart type="given">Frederikus</namePart>
<namePart type="family">Hudi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nina</namePart>
<namePart type="family">Nanakorn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Justin</namePart>
<namePart type="family">Vasselli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yusuke</namePart>
<namePart type="family">Sakai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hidetaka</namePart>
<namePart type="family">Kamigaito</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akihiro</namePart>
<namePart type="family">Tamura</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Taro</namePart>
<namePart type="family">Watanabe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-444-6</identifier>
</relatedItem>
<abstract>Existing machine translation and multimodal translation benchmarks mainly focus on text-only inputs or short examples with tightly aligned image-text pairs. Academic lectures present a different challenge: spoken explanations, presentation slides, and domain-specific terminology interact across long-form discourse. LecTrans is constructed from 383 hours of expert-taught online lectures and provides transcripts, presentation slides, and an eight-language parallel evaluation subset covering English and seven target languages. The 162-minute translation subset spans 17 subjects and contains human-corrected English transcripts, verified slide text, and human-verified translations of both transcripts and slides. We define lecture translation as two complementary tasks: transcript translation for spoken lecture content and slide translation for visually presented text. We report text-only and multimodal baselines, showing that noisy transcripts, loosely aligned slide context, and domain-specific content remain challenging for current systems. LecTrans provides a testbed for evaluating multimodal translation in long-form academic lectures.</abstract>
<identifier type="citekey">hudi-etal-2026-lectrans</identifier>
<location>
<url>https://aclanthology.org/2026.findings-aacl.93/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>1872</start>
<end>1897</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T LecTrans: A Multimodal Translation Benchmark for Academic Lectures
%A Hudi, Frederikus
%A Nanakorn, Nina
%A Vasselli, Justin
%A Sakai, Yusuke
%A Kamigaito, Hidetaka
%A Tamura, Akihiro
%A Watanabe, Taro
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F hudi-etal-2026-lectrans
%X Existing machine translation and multimodal translation benchmarks mainly focus on text-only inputs or short examples with tightly aligned image-text pairs. Academic lectures present a different challenge: spoken explanations, presentation slides, and domain-specific terminology interact across long-form discourse. LecTrans is constructed from 383 hours of expert-taught online lectures and provides transcripts, presentation slides, and an eight-language parallel evaluation subset covering English and seven target languages. The 162-minute translation subset spans 17 subjects and contains human-corrected English transcripts, verified slide text, and human-verified translations of both transcripts and slides. We define lecture translation as two complementary tasks: transcript translation for spoken lecture content and slide translation for visually presented text. We report text-only and multimodal baselines, showing that noisy transcripts, loosely aligned slide context, and domain-specific content remain challenging for current systems. LecTrans provides a testbed for evaluating multimodal translation in long-form academic lectures.
%U https://aclanthology.org/2026.findings-aacl.93/
%P 1872-1897
Markdown (Informal)
[LecTrans: A Multimodal Translation Benchmark for Academic Lectures](https://aclanthology.org/2026.findings-aacl.93/) (Hudi et al., Findings 2026)
ACL
- Frederikus Hudi, Nina Nanakorn, Justin Vasselli, Yusuke Sakai, Hidetaka Kamigaito, Akihiro Tamura, and Taro Watanabe. 2026. LecTrans: A Multimodal Translation Benchmark for Academic Lectures. In Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026, pages 1872–1897, Hengqin, China. Association for Computational Linguistics.