MateCat worked at a new frontier of Computer Assisted Translation (CAT) technology, that is, how to effectively and ergonomically integrate Machine Translation (MT) within the human translation workflow. At a time in which MT was mainly trained with the objective of creating the most comprehensible output, MateCat targeted the development of MT technology aimed at minimizing the translatorโs post-edit effort. To this end, MateCat developed an enhanced web-based CAT tool offering new MT capabilities, such as automatic adaption to the translated content, online learning from user corrections, and automatic quality estimation.
๐๐ผ Excited to share our work on Speech Foundation Model for data crowdsourcing at COLING 2025 ๐๐ผ
Our co-author Laurent Besacier (@laurent_besacie) at NAVER LABS Europe will be presenting -- don't miss it.
๐๐ผ Details: https://mt.fbk.eu/1-paper-accepted-at-coling-2025
Exciting news: @iwslt is co-located with #ACL2025NLP again this year! ๐
Interested in speech processing? Check out the new task on instruction following โ any model can participate! ๐
๐
Data release: April 1
โณ Submission deadline: April 15
Donโt miss it! ๐ฌ #NLP #SpeechTech
Weekly pick from the #MeetweenScientificWatch: โVideo-SALMONN: Speech-enhanced audio-visual large language modelsโ โ Redefining video comprehension with speech-aware AV-LLMs and groundbreaking QA accuracy. ๐ฅ๐ค๐ค
Iโm glad to announce that our work โHow "Real" is Your Real-Time Simultaneous Speech-to-Text Translation System?โ has been accepted at the Transactions of @aclanthology (TACL)! ๐
The preprint is available here: