Meetween

Meetween

IT Services and IT Consulting

EU-funded AI partnership. Avatars for the future of videoconferencing. Foundation models for speech.

About us

Industry
IT Services and IT Consulting
Company size
11-50 employees
Type
Partnership
Founded
2024

Updates

  • Meetween reposted this

    View organization page for HE-UTTER, graphic

    84 followers

    🔶 UTTER 2nd User Day: Agenda What´s in store for our 2nd User Day? We have an exciting agenda lined up! Vlad Niculae will kick off the event with a deep dive into the UTTER project. Next, Ben Peters will cover robustness in translation models in our Reserach Spotlight, followed by Marcely Zanon Boito, who will present mHuBERT-147, a compact and powerful multilingual speech presentation model. Sébastien Bratières will also introduce Meetween: My Personal AI Mediator for Virtual Meetings, a project with the mission to build the AI-based technology solutions needed to power the next generation of video conferencing platforms to support smooth, engaging and barrier-free collaboration across languages, geographies and time zones. This event is fully remote and open to the public and we invite everyone interested in multilingual and multimodal LLMs and NLP research to join us! 👉 Stay tuned for more insights on the full agenda. Click here to register:https://lnkd.in/d3whia_8 #UTTERUserDay #XR #airesearch #NLP #FSTP #horizoneurope #multimodalLLMs #paneldiscussion

    • No alternative text description for this image
  • We're proud to announce the launch of Meetween, a groundbreaking initiative funded by the European Union’s Horizon Europe program with a budget of €7.2M to develop the science-based technology solutions needed for next-generation #videoconferencing. The project will be carried out by a consortium led by Translated, a leading provider of AI-powered language solutions, and including Fondazione Bruno Kessler - FBK, Karlsruhe Institute of Technology (KIT), ITU, TAUS, Zoom, Academic Computer Centre CYFRONET AGH, and Pi School. Meetween aims to overcome the barriers and separation caused by #language, culture, physical distance, time zones, scheduling conflicts, and distractions. We will leverage foundation models for speech with their rich, integrated algorithmic capabilities and self-supervised training on large datasets. Our avatars, assistants, and support technology will adapt to participants' contextual, cultural, and regional idiosyncrasies, including linguistic ones. This will allow us to promote and facilitate business collaboration across the European Union by providing real-time, machine-learning-powered speech-to-speech translation, summarization, and virtual assistant services for online meetings. Meetween's project outputs, including datasets, neural network models, and evaluation benchmark platforms, will be made open source. Follow Meetween for updates. #AI #ML #innovation #L10N

Similar pages