
Institute · Trento, Italy
FBK Machine Translation Unit
The machine translation group at Fondazione Bruno Kessler, publishing open speech and text translation datasets and software.
Trento, Italy
- Type
- Institute
- Based in
- Trento, Italy
- Key papers
- 4
- Featured projects
- 8
About
The MT group at Fondazione Bruno Kessler works on speech and text technologies for multilingual communication. Its research covers neural architectures and multimodal resources for applications such as subtitling, interpreting and gender inclusivity, and it sits within FBK's Digital Industry center.
The group's site lists corpora including MCIF, mGeNTE, MOSEL, Speech-MASSIVE and WAGS, and software such as simulstream. Researchers named on the site include Luisa Bentivogli and Beatrice Savoldi.
Mission and ambitionsThe unit aims to advance speech and text translation, with attention to inclusivity, and to share its data and code openly.
Known for
- 01
Open speech translation corpora
- 02
Simultaneous translation software
- 03
Gender-fair machine translation research
Milestones
- 2026
simulstream release
Software for simultaneous speech translation released.
Open source
Collection of open source speech data (166 stars)
Open source code of works from the FBK MT unit (60 stars)
Translation Memory Open-source Purifier (35 stars)
Library for simultaneous speech recognition and translation (33 stars)
- FBK-MT/moseldataset
MOSEL speech data collection, the unit's most downloaded repository.
- FBK-MT/Speech-MASSIVEdataset
Multilingual spoken language understanding dataset.
- FBK-MT/MCIFdataset
Crosslingual instruction-following benchmark.
- FBK-MT/fama-mediummodel
Open-science speech foundation model.
Research papers
Notable releases
Openness
Code and datasets are public on GitHub and Hugging Face under Apache-2.0 for several repositories.
People
- Luisa BentivogliHead of Unit
- Matteo NegriSenior Research Scientist
- Mauro CettoloSenior Research Scientist
- Marco GaidoResearcher
- Beatrice SavoldiResearcher, gender bias in NLP
Listed from the lab's own pages and public profiles.
Contact and community
The unit's website blocked automated access during checking; use the GitHub (hlt-mt) or Hugging Face pages for project contact.
Sources
Compiled from the lab's public pages, Hugging Face and GitHub. Last checked 2026-11-17. Something wrong or out of date? Tell us on Discord.



