My Personal AI Mediator for Virtual MEETtings BetWEEN People

Informations projet

Meetween

N° de convention de subvention: 101135798

DOI

10.3030/101135798

Date de signature de la CE 3 Novembre 2023

Date de début 1 Janvier 2024

Date de fin 31 Decembre 2027

Financé au titre de

• Digital, Industry and Space

Coût total

€ 7 239 725,00

Contribution de l’UE

€ 7 239 725,00

Coordonné par

TRANSLATED SRL
Italy

CORDIS fournit des liens vers les livrables publics et les publications des projets HORIZON.

Les liens vers les livrables et les publications des projets du 7e PC, ainsi que les liens vers certains types de résultats spécifiques tels que les jeux de données et les logiciels, sont récupérés dynamiquement sur OpenAIRE .

Livrables

Other (5)

Downstream task trained models - V1

This deliverable reports the efforts to adapt the best models to satisfy production constraints, such as low latency, low flickering, and low computational costs, and develop APIs to serve different applications and needs.

Speech LMM open release - V1

SpeechLMM is released three times during the project and each release is documented in a version of this deliverable.

Model cards - V1

All partners will author a deliverable containing model cards.

Speech LMM training journal - V1

This deliverable documents the “making of” SpeechLMM and is used for dissemination to the scientific and technology community in the form of the SpeechLMM training journal.

Human factors and security scenarios - V1

This deliverable provides human factors and a security report describing vulnerability and criticality scenarios.

Data sets, microdata, etc (1)

Mumospee open release - V1

This deliverable documents the creation of one of the largest open multimodal datasets, Mumospee, released over three generations, in coordination with the corresponding SpeechLMM generation activities.

Publications

Conference proceedings (45)

Factorized-VITS: Decoupling Prosody and Text in End-to-End Speech Synthesis without External or Secondary Aligner

Auteurs: Yining Liu, Alexander Waibel
Publié dans: ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2025
Éditeur: IEEE
DOI: 10.1109/ICASSP49660.2025.10890003

SimulSeamless: FBK at IWSLT 2024 Simultaneous Speech Translation

Auteurs: Sara Papi, Marco Gaido, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 21st International Conference on Spoken Language Translation (IWSLT 2024), 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.IWSLT-1.11

Weight Factorization and Centralization for Continual Learning in Speech Recognition

Auteurs: Enes Yavuz Ugan*, Ngoc-Quan Pham*, Alexander Waibel
Éditeur: INTERSPEECH 2025
DOI: 10.48550/ARXIV.2506.16574

The Warmup Dilemma: How Learning Rate Strategies Impact Speech-to-Text Model Convergence

Auteurs: Marco Gaido*, Sara Papi*, Luisa Bentivogli, Alessio Brutti, Mauro Cettolo, Roberto Gretter, Marco Matassoni, Mohamed Nabih, Matteo Negri.
Éditeur: IWSLT2025
DOI: 10.48550/ARXIV.2505.23420

NUTSHELL: A Dataset for Abstract Generation from Scientific Talks

Auteurs: Maike Züfle, Sara Papi, Beatrice Savoldi, Marco Gaido, Luisa Bentivogli, Jan Niehues
Publié dans: Proceedings of the 22nd International Conference on Spoken Language Translation (IWSLT 2025), 2025
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2025.IWSLT-1.2

MSA-ASR: Efficient Multilingual Speaker Attribution with frozen ASR Models

Auteurs: Thai-Binh Nguyen, Alexander Waibel
Éditeur: ICASSP 2025
DOI: 10.48550/ARXIV.2411.18152

A Bayesian Optimization Approach to Machine Translation Reranking

Auteurs: Julius Cheng, Maike Züfle, Vilém Zouhar, Andreas Vlachos
Éditeur: NAACL 2025 Main
DOI: 10.48550/ARXIV.2411.09694

FBK@IWSLT Test Suites Task: Gender Bias evaluation with MuST-SHE

Auteurs: Beatrice Savoldi, Marco Gaido, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 21st International Conference on Spoken Language Translation (IWSLT 2024), 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.IWSLT-1.10

MOSEL: 950,000 Hours of Speech Data for Open-Source Speech Foundation Model Training on EU Languages

Auteurs: Marco Gaido, Sara Papi, Luisa Bentivogli, Alessio Brutti, Mauro Cettolo, Roberto Gretter, Marco Matassoni, Mohamed Nabih, Matteo Negri
Publié dans: Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.EMNLP-MAIN.771

StreamAtt: Direct Streaming Speech-to-Text Translation with Attention-based Audio History Selection

Auteurs: Sara Papi, Marco Gaido, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.ACL-LONG.202

Quality Estimation with $k$-nearest Neighbors and Automatic Evaluation for Model-specific Quality Estimation

Auteurs: Dinh, Tu Anh; Palzer, Tobias; Niehues, Jan
Publié dans: 2024
Éditeur: EAMT 2024
DOI: 10.5445/IR/1000174743

GFG - Gender-Fair Generation: A CALAMITA Challenge

Auteurs: Simona Frenda, Andrea Piergentili, Beatrice Savoldi, Marco Madeddu, Martina Rosola, Silvia Casola, Chiara Ferrando, Viviana Patti, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 10th Italian Conference on Computational Linguistics (CLiC-it 2024), 2024
Éditeur: CEUR Workshop Proceedings

FINDINGS OF THE IWSLT 2024 EVALUATION CAMPAIGN

Auteurs: Ibrahim Said Ahmad, Antonios Anastasopoulos, Ondřej Bojar, Claudia Borg, Marine Carpuat, Roldano Cattoni, Mauro Cettolo, William Chen, Qianqian Dong, Marcello Federico, Barry Haddow, Dávid Javorský, Mateusz Krubiński, Tsz Kim Lam, Xutai Ma, Prashant Mathur, Evgeny Matusov, Chandresh Maurya, John McCrae, Kenton Murray, Satoshi Nakamura, Matteo Negri, Jan Niehues, Xing Niu, Atul Kr. Ojha, John Ortega, Sara Papi, Peter Polák, Adam Pospíšil, Pavel Pecina, Elizabeth Salesky, Nivedita Sethiya, Balaram Sarkar, Jiatong Shi, Claytone Sikasote, Matthias Sperber, Sebastian Stüker, Katsuhito Sudoh, Brian Thompson, Alex Waibel, Shinji Watanabe, Patrick Wilken, Petr Zemánek, Rodolfo Zevallos
Publié dans: Proceedings of the 21st International Conference on Spoken Language Translation (IWSLT 2024), 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.IWSLT-1.1

Speech Translation with Speech Foundation Models and Large Language Models: What is There and What is Missing?

Auteurs: Marco Gaido, Sara Papi, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.ACL-LONG.789

Contrastive Learning for Task-Independent SpeechLLM-Pretraining

Auteurs: Maike Züfle, Jan Niehues
Éditeur: ACL 2025 Findings
DOI: 10.48550/ARXIV.2412.15712

Translation in the Hands of Many:Centering Lay Users in Machine Translation Interactions

Auteurs: Beatrice Savoldi, Alan Ramponi, Matteo Negri, Luisa Bentivogli.
Éditeur: arXiv
DOI: 10.48550/ARXIV.2502.13780

Decoupled Vocabulary Learning Enables Zero-Shot Translation from Unseen Languages

Auteurs: Mullov, Carlos; Pham, Quan; Waibel, Alexander
Publié dans: Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2024
Éditeur: ACL
DOI: 10.5445/IR/1000174872

KIT's Offline Speech Translation and Instruction Following Submission for IWSLT 2025

Auteurs: Sai Koneru*, Maike Züfle*, Thai-Binh Nguyen, Seymanur Akti, Jan Niehues, Alexander Waibel
Éditeur: IWSLT 2025
DOI: 10.48550/ARXIV.2505.13036

Findings of the IWSLT 2025 Evaluation Campaign

Auteurs: Idris Abdulmumin, Victor Agostinelli, Tanel Alumäe, Antonios Anastasopoulos, Luisa Bentivogli, Ondřej Bojar, Claudia Borg, Fethi Bougares, Roldano Cattoni, Mauro Cettolo, Lizhong Chen, William Chen, Raj Dabre, Yannick Estève, Marcello Federico, Mark Fishe
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2025.IWSLT-1.44

Optimizing Rare Word Accuracy in Direct Speech Translation with a Retrieval-and-Demonstration Approach

Auteurs: Li, Siqi; Liu, Danni; Niehues, Jan
Publié dans: Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, 2024
Éditeur: EMNLP
DOI: 10.48550/ARXIV.2409.09009

Towards Better Disentanglement in Non-Autoregressive Zero-Shot Expressive Voice Conversion Voice Conversion

Auteurs: Seymanur Akti, Tuan Nam Nguyen, Alexander Waibel
Éditeur: INTERSPEECH 2025
DOI: 10.48550/ARXIV.2506.04013

Cocktail-Party Audio-Visual Speech Recognition

Auteurs: Thai-Binh Nguyen, Ngoc-Quan Pham, Alexander Waibel
Éditeur: INTERSPEECH 2025
DOI: 10.48550/ARXIV.2506.02178

MAGNET - MAchines GeNErating Translations: A CALAMITA Challenge

Auteurs: Mauro Cettolo, Andrea Piergentili, Sara Papi, Marco Gaido, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 10th Italian Conference on Computational Linguistics (CLiC-it 2024), 2024
Éditeur: CEUR Workshop Proceedings

An LLM-as-a-judge Approach for Scalable Gender-Neutral Translation Evaluation

Auteurs: Andrea Piergentili, Beatrice Savoldi, Matteo Negri, Luisa Bentivogli
Publié dans: Proceedings of the 3rd Workshop on Gender-Inclusive Translation Technologies (GITT 2025), 2025
Éditeur: European Association for Machine Translation

From Speech to Summary: A Comprehensive Survey of Speech Summarization

Auteurs: Retkowski, Fabian; Züfle, Maike; Sudmann, Andreas; Pfau, Dinah; Niehues, Jan; Waibel, Alexander
Publié dans: 2025
Éditeur: EMNLP 2025 Main
DOI: 10.5445/IR/1000180972

Recent Highlights in Multilingual and Multimodal Speech Translation

Auteurs: Danni Liu and Jan Niehues
Éditeur: IWSLT 2024
DOI: 10.48550/ARXIV.2506.11160

Impact of Face Alignment on Face Image Quality

Auteurs: Eren Onaran, Erdi Sarıtaş, Hazım Kemal Ekenel
Publié dans: EAI ROSENET 2024
Éditeur: EAI ROSENET 2024
DOI: 10.48550/ARXIV.2507.20953

Speech Foundation Models and Crowdsourcing for Efficient, High-Quality Data Collection

Auteurs: Beomseok Lee, Marco Gaido, Ioan Calapodescu, Laurent Besacier, Matteo Negri
Publié dans: Proceedings of the 31st International Conference on Computational Linguistics, 2025
Éditeur: Association for Computational Linguistics

PIER: A Novel Metric for Evaluating What Matters in Code-Switching

Auteurs: Enes Yavuz Ugan, Ngoc-Quan Pham, Leonard Bärmann, and Alex Waibel
Éditeur: ICASSP 2025
DOI: 10.48550/ARXIV.2501.09512

FAMA: The First Large-Scale Open-Science Speech Foundation Model for Italian and English

Auteurs: Sara Papi*, Marco Gaido*, Luisa Bentivogli, Alessio Brutti, Mauro Cettolo, Roberto Gretter, Marco Matassoni, Mohamed Nabih, Matteo Negri.
Éditeur: CLIC-it 2025
DOI: 10.48550/ARXIV.2505.22759

Prepending or Cross-Attention for Speech-to-Text? An Empirical Comparison

Auteurs: Tsz Kin Lam, Marco Gaido, Sara Papi, Luisa Bentivogli, Barry Haddow
Publié dans: Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), 2025
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2025.NAACL-LONG.153

Continuously Learning New Words in Automatic Speech Recognition

Auteurs: Christian Huber, Alexander Waibel
Éditeur: ICASSP 2025
DOI: 10.48550/ARXIV.2401.04482

Cross-lingual Evaluation of Multilingual Text Generation

Auteurs: Shamil Chollampatt, Minh Quang Pham, Sathish Reddy Indurthi, Marco Turchi
Publié dans: Proceedings of the 31st International Conference on Computational Linguistics, 2025
Éditeur: Association for Computational Linguistics

Audio-driven Talking Face Generation with Stabilized Synchronization Loss

Auteurs: Dogucan Yaman, Fevziye Irem Eyiokur, et al.
Éditeur: EMNLP 2025 Main
DOI: 10.48550/ARXIV.2307.09368

Analyzing the Feature Extractor Networks for Face Image Synthesis

Auteurs: Erdi Sarıtaş and Hazım Kemal Ekenel
Éditeur: IEEE FG 2024
DOI: 10.48550/ARXIV.2406.02153

What the Harm? Quantifying the Tangible Impact of Gender Bias in Machine Translation with a Human-centered Study

Auteurs: Beatrice Savoldi, Sara Papi, Matteo Negri, Ana Guerberof-Arenas, Luisa Bentivogli
Publié dans: Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.EMNLP-MAIN.1002

Explainability for Speech Models: On the Challenges of Acoustic Feature Selection

Auteurs: Dennis Fucci, Beatrice Savoldi, Marco Gaido, Matteo Negri, Mauro Cettolo, Luisa Bentivogli
Publié dans: Proceedings of the 10th Italian Conference on Computational Linguistics (CLiC-it 2024), 2024
Éditeur: CEUR Workshop Proceedings

Blending LLMs into Cascaded Speech Translation: KIT's Offline Speech Translation System for IWSLT 2024

Auteurs: Sai Koneru, Thai-Binh Nguyen, Ngoc-Quan Pham, Danni Liu, Zhaolin Li, Alexander Waibel, Jan Niehues
Éditeur: IWSLT 2024
DOI: 10.48550/ARXIV.2406.16777

Assessing the Use of Face Swapping Methods as Face Anonymizers in Videos

Auteurs: Mustafa İzzet Muştu, Hazım Kemal Ekenel
Éditeur: DSP 2025
DOI: 10.48550/ARXIV.2505.20985

Different Speech Translation Models Encode and Translate Speaker Gender Differently

Auteurs: Dennis Fucci, Marco Gaido, Matteo Negri, Luisa Bentivogli, André Martins, Giuseppe Attanasio
Éditeur: ACL 2025 Main
DOI: 10.48550/ARXIV.2506.02172

How do Multimodal Foundation Models Encode Text and Speech? An Analysis of Cross-Lingual and Cross-Modal Representations

Auteurs: Hyunji Lee, Danni Liu, Supriti Sinhamahapatra, Jan Niehues
Éditeur: NAACL 2025
DOI: 10.48550/ARXIV.2411.17666

Streaming Non-Autoregressive Model for Accent Conversion and Pronunciation Improvement

Auteurs: Tuan-Nam Nguyen, Ngoc-Quan Pham, Seymanur Akti, Alexander Waibel
Éditeur: INTERSPEECH 2025
DOI: 10.48550/ARXIV.2506.16580

How to Connect Speech Foundation Models and Large Language Models? What Matters and What Does Not

Auteurs: Francesco Verdini, Pierfrancesco Melucci, Stefano Perna, Francesco Cariaggi, et al.
Éditeur: INTERSPEECH 2025
DOI: 10.48550/ARXIV.2409.17044

Findings of the Quality Estimation Shared Task at WMT 2024: Are LLMs Closing the Gap in QE?

Auteurs: Chrysoula Zerva, Frederic Blain, José G. C. De Souza, Diptesh Kanojia, Sourabh Deoghare, Nuno M. Guerreiro, Giuseppe Attanasio, Ricardo Rei, Constantin Orasan, Matteo Negri, Marco Turchi, Rajen Chatterjee, Pushpak Bhattacharyya, Markus Freitag, André Martins
Publié dans: Proceedings of the Ninth Conference on Machine Translation, 2024
Éditeur: Association for Computational Linguistics
DOI: 10.18653/V1/2024.WMT-1.3

Audio-Visual Speech Representation Expert for Enhanced Talking Face Video Generation and Evaluation

Auteurs: Dogucan Yaman, Fevziye Irem Eyiokur, et al.
Éditeur: CVPRW 2024
DOI: 10.48550/ARXIV.2405.04327

Peer reviewed articles (2)

How “Real” is Your Real-Time Simultaneous Speech-to-Text Translation System?

Auteurs: Sara Papi, Peter Polák, Dominik Macháček, Ondřej Bojar
Publié dans: Transactions of the Association for Computational Linguistics, Numéro 13, 2025, ISSN 2307-387X
Éditeur: MIT Press
DOI: 10.1162/TACL_A_00740

A decade of gender bias in machine translation

Auteurs: Beatrice Savoldi, Jasmijn Bastings, Luisa Bentivogli, Eva Vanmassenhove
Publié dans: Patterns, Numéro 6, 2025, ISSN 2666-3899
Éditeur: Elsevier BV
DOI: 10.1016/J.PATTER.2025.101257

Other (2)

Facial Attribute Based Text Guided Face Anonymization

Auteurs: Muştu, Mustafa İzzet; Ekenel, Hazım Kemal
Publié dans: 2025
Éditeur: arXiv
DOI: 10.48550/ARXIV.2505.21002

Early-Exit and Instant Confidence Translation Quality Estimation

Auteurs: Vilém Zouhar, Maike Züfle, Beni Egressy, Julius Cheng, Jan Niehues
Éditeur: arXiv
DOI: 10.48550/ARXIV.2502.14429