EMHCAM · Embodied Multimodal Human Communication and Modelling
„Хоризонт Европа“ — Действия „Мария Склодовска-Кюри“
- Период
- 2027-01-01 → 2030-12-31
- Финансиране от ЕС
- 4 711 376 €
- Участници
- 18
- Схема
- HORIZON-TMA-MSCA-DN
Линиите свързват координатора с партньорите. За проекти отпреди 2014 г. CORDIS не винаги дава точни координати. Тези точки са на ниво град или държава.
Накратко на български
Връзката между речта и жестовете при човешката комуникация се анализира чрез събиране на аудио-визуални данни и създаване на компютърни модели. Това помага за подобряване на слуховите апарати, технологиите за учене и социалните роботи в здравеопазването.
Кратко обяснение, генерирано от езиков модел по текста на CORDIS. Оригиналът е по-долу.
Цел на проекта
The EMHCAM Doctoral Network will train 15 doctoral candidates to advance European research in Embodied Multimodal Human Communication And Modelling. Despite rapid progress in language and vision AI, current systems lack grounding in the embodied and context-dependent nature of human communication. This limits progress in key application domains such as socially interactive agents, accessibility technologies and health-related human–machine interaction. EMHCAM addresses this gap by integrating three research communities — gesture and cognitive interaction studies, multimodal corpus development and multimodal data science — into a single, coherent doctoral training network.The network tackles three interconnected scientific challenges: 1) developing new theoretical accounts of how speech and gesture jointly structure communication across languages, speaker profiles and interaction settings; 2) creating the next generation of embodied multimodal corpora, linking audio–visual and motion-captured data with shared annotation conventions and making them available through ethical data-sharing methods; 3) designing computational models that analyse and generate multimodal behaviour, including LLM-based generation pipelines and context-aware behavioural modelling.The resulting resources — open multimodal datasets, validated annotation schemes, evaluation benchmarks and ethically robust modelling pipelines — will have concrete impact across academia and industry. Application areas include intelligent hearing aid design, socially interactive agents for mental healthcare and multimodal learning technologies. By combining advanced technical training, strong industry participation and an interdisciplinary, cross-European research environment, EMHCAM will produce a new generation of researchers equipped to shape human-centred, trustworthy and inclusive multimodal technologies in line with European strategic priorities.
Оригинален текст от CORDIS (на английски).
Участници
- KOBENHAVNS UNIVERSITET · KOBENHAVNКоординаторДания
- Aryza Ireland Limited · DublinНиво градИрландия
- ENDURAE VOICE TECHNOLOGY OU · TallinnНиво градЕстония
- JOHANN WOLFGANG GOETHE-UNIVERSITAET FRANKFURT AM MAIN · Frankfurt Am MainГермания
- KATHOLIEKE UNIVERSITEIT LEUVEN · LeuvenБелгия
- KUNGLIGA TEKNISKA HOEGSKOLAN · StockholmШвеция
- LINNEUNIVERSITETET · VaxjoШвеция
- LUNDS UNIVERSITET · LundШвеция
- SORBONNE UNIVERSITE · ParisФранция
- STOCKHOLMS UNIVERSITET · StockholmШвеция
- THE PROVOST, FELLOWS, FOUNDATION SCHOLARS & THE OTHER MEMBERS OF BOARD, OF THE COLLEGE OF THE HOLY & UNDIVIDED TRINITY OF QUEEN ELIZABETH NEAR DUBLIN · DublinИрландия
- TOBII AB · DanderydШвеция
- UNIVERSIDAD POMPEU FABRA · BarcelonaИспания
- UNIVERSITAET BIELEFELD · BielefeldГермания
- UNIVERSITY COLLEGE LONDON · LondonОбединеното кралство
- Université de Montpellier Paul-Valéry · MontpellierФранция
- WS Audiology A/S · LyngeНиво градДания
- iMotions A/S · CopenhagenДания
Връзки
Данни: CORDIS, © Европейски съюз