The end of the containment allowed social and professional activities resumption while raising questions about the risk of SARS-CoV-2 spreading and protective measures to involve. Conversations (normal speech) or professional interactions (loud speech) have been identified as an increased risk of SARS-CoV-2 exposure related to high production of droplets in the expired air. Potential COVID-19 transmission by speech is an issue frequently raised in medias highlighting the imprecise and contradictory data emerging in literature. Here, we propose a systematic study analyzing the velocity of droplets in the expired air during different speech conditions (normal speech, projected, interacting). This will allow to accurately define droplets dispersion and provide objective data regarding the potential risk of contamination to health authorities. Our aim is the SYSTEMATIC STUDY OF DROPLETS PRODUCED DURING SPEECH ACCORDING TO THE VOICE INTENSITY AND TO THE PHONETIC CONTENTS (VOWELS, CONSONANT, INTONATION). This proposal is part of the theme "PREVENTION AND CONTROL OF INFECTION" and more particularly in "STUDY OF HUMAN AND ENVIRONMENTAL FACTORS INFLUENCING TRANSMISSION". This study is designed to provide reference data on droplets dissemination during speech, taking into account sound characteristics and speech conditions. Indeed, these results will be a strong base of knowledge for the implementation of protective measures in various fields: adaptation of masks, validation of distance measures in education and more generally professional environments. To achieve this goal, the interdisciplinary and complementary team (Laboratoire Parole et Langage LPL specialized in phonetic sciences and Institut Universitaire des Sciences et Techniques de l’Ingénieur IUSTI specialized in fluid mechanics) involved in this project will have to tackle with the following challenging questions: 1) What is the velocity of the expired air during speech? We propose a systematic analysis of this velocity with a hot-wire flowmeter during production of a linguistic corpus exploring all the items of articulation of speech (in French) by a large cohort of 50 speakers 2) What is the number of droplets in the expired air during speech? We propose to use a Particle Analyser to obtain a kind of size spectrography of emitted particles, knowing the fact that trajectories of droplets is a function of the size/weight 3) What is the spatial dissemination of droplets in the expired air during speech? We will use high speed cameras to analyze velocity and trajectory of the cloud of particles emitted during speech following a preliminary study performed during the confinement (1) 4) At the end of the experimentations, all data will be entered and will allow Computational Fluid Dynamic (CFD) studies analyzing the transport, dispersion and evaporation of droplets emitted according to different speech conditions. Then, in 9 months, we will be able to share with government agencies responsible for health instructions, and more generally the scientific community, our data concerning the different speaking situations, especially for loud voicing (education, professional). This will help to improve the security of the population and fall under the prevention component of the RA-COVID call for tenders. (1) Giovanni A, Radulesco T, Bouchet G, Mattei A, Revis J, Bogdanski E, Michel J. Transmission of droplet-conveyed infectious agents such as Sars Cov2 by speech and vocal exercises during speech therapy : preliminary experiment concerning airflow velocity. Europ Arch ORL 2020 20-01484R1 published on line July 2020

Current European language education policies are governed by two pillars: the task-based approach and verbal interaction. Since the mid-1990s, with the emergence of computer-mediated communication tools, there has been an introduction of telecollaboration and virtual exchange in foreign language learning, with a progressive institutionalization of these forms of learning arriving at the Erasmus + Virtual Exchange project of the European Commission. The scientific goal of our network is to advance our knowledge of foreign language acquisition in computer-mediated communication and to translate this knowledge into the design of language learning environments and tasks in the private and public sectors. This scientific goal is pursued by a network currently comprising eight European partners in six countries, two non-European partners and one partner from the United Kingdom. The disciplines present are language sciences, education sciences and computer science. The areas of expertise in the consortium are multimodal analysis of the interactive dynamics underlying language acquisition (conversational alignments, specific phenomena such as negotiations of meaning), multimodal corpus building and analysis using (semi-)automatic natural language processing tools, task design and digital learning environments, online language training management and marketing. This consortium is expected to expand with greater participation from the private sector. The consortium will pursue R&D by training and collaborating with a group of ten young researchers. These young researchers will be trained to develop transversal skills that will enable them to navigate between the public and private sectors. This training will be done by combining distance learning, training with periods of mobility (between academic partners and in companies), and an annual one-week training course bringing together all the partners. Within the framework of the MRSEI project, the following tasks will be carried out: formalization of the scientific programme for submission to the AAP MSCA ITN; recruitment of non-academic partners; formalization of the training of young researchers; planning of dissemination; writing of the ITN project. These tasks will be accomplished through two three-day meetings with the entire consortium and close collaboration between the work-package managers of the future ITN project. The consortium will be helped by Protisvalor, the subsidiary of Aix Marseille for the support of researchers in the writing of European projects. By supporting the future ITN project by a French team, this project will consolidate the role of France and the tools and infrastructures it finances (such as the Ortolang platform, Equipex ANR-11-EQPX-0032) in European research.

During verbal communication individuals tend to align at different linguistic levels. Similarly, playing music requires fine temporal alignment and coordination abilities, in order to anticipate and adapt to other’s actions in real time. The aim of the project is to study the effect of musical rhythm on interpersonal verbal coordination in adults, children and children with hearing impairment. We will first quantify with behavioral and neural measures the interpersonal coordination during verbal exchanges. Then, we will test to what extent and how music training modifies these coordination metrics. The originality of the project lies in the use of a multilevel approach to conversation, front edge analysis techniques on speech and neural data. Importantly, this is the first attempt to study the effect of rhythmic joint-action (music) on different levels of language coordination in hearing impaired children.

In a conversation, feedback is mostly performed through short utterances produced by another participant than the main current speaker. These utterances are among the most frequent in conversational data. They are also considered as crucial communicative tools for achieving coordination in dialogue. They have been the topic of various descriptive studies and often given a central role in applications such as dialogue systems. The present project addresses this issue from a linguistic viewpoint and combines fine-grained corpus analyses of semi-controlled data with formal and statistical modeling. At the formal level, the dynamic turn of semantics has laid the ground for rich formal approaches of discourse in which the semantic/pragmatic interface can play its role. Such models allow for a rich yet precise characterization of dialogue communicative functions. The impoverished aspect of the linguistic material in these utterances allows for a truly multi-dimensional analysis that can unveil how different linguistic domains (morpho-syntax, prosody, visual channel) combine to convey meaning and achieve communicative goals. Statistical model will go in hand with the formal model giving it more empirical support and allowing it to focus on truly representative and reliable phenomena. The statistical model will be adapted to create a classification system determining the main communicative functions of a feedback item based on its properties on the different observable dimensions and contextual information. Human-systems interfaces as well systems processing audio and video data have reached interesting achievements in the domain of feedback but is waiting for more linguistic models to continue to improve. A better understanding of the form/function relation of feedback can help extracting a wide range of elements from conversational data such as decisions taken, information accepted, opinion changes or humor-involving sequences.

It is becoming increasingly realistic to exploit transcriptions of spoken data for tasks that require comprehension of what is said in a conversation. SUMM-RE will combine expertise in theories of discourse interpretation with recent developments in distant supervision to improve the automatic production of meeting summaries and minutes from spoken data. State of the art approaches to abstractive summarization treat discourse as a mere linear sequence of utterances. SUMM-RE posits that by exploiting information about discourse relations and the rich structures determined by relations between utterances, we can significantly improve models for abstractive summarization. A major hurdle to developing more sophisticated models of discourse structure for spoken, multiparty conversation is a lack of appropriate training data. SUMM-RE will address this problem in two ways. First, it will create a new and unique corpus of meeting-like interactions in French. Second, it will label this corpus and a large corpus of meeting-like interactions in English for discourse structure. The annotation approach will extend recent developments in distant supervision to develop labelling functions that can be used to automatically label large amounts of data. This approach has the very attractive advantage of harnessing linguistic expertise while keeping manual annotation to a minimum. The automatically annotated data will be used to improve algorithms for both short topic summaries and more detailed meeting minutes. These algorithms in turn will be integrated into the lead partner's (LINAGORA's) semi-automatic summarization tool to significantly improve the output for its users. All project results (corpus and algorithms) will be released under an open-source license as a part of LINAGORA's LinTo/Conversation Manager offer.