VoxCrim deals with speaker authentication and more precisely forensic voice comparison. VoxCrim targets national security and legal/justice application contexts. The project's aims correspond to "ANR Défi 9". VoxCrim proposes a validated scientific objective framework available for all kinds of forensic voice comparison methods (automatic or phonetic). The goal is to develop certified standards to delimit the specific areas where voice comparison methods are applicable. VoxCrim gathers computer scientists (LIA) with phoneticians (LPL, LPP), experts in standardization (LNE) and members of the forensic department of Police Nationale (SDPTS). Members of the forensics department of Gendarmerie Nationale will take part in the project as a center of expertise. VoxCrim's aims are organised along two time scales: (1) analyses of well-controlled recordings in the near future, and (2) an enlarged area of application later on. (1) VoxCrim proposes to develop methods and databases based on the results of ANR Fabiole in order to extend the innovative concept of "box rule" (i.e. a set of conditions where voice comparison is perfectly applicable and where certification is possible). (2) Then, the project aims to add dimensions to the standards of voice comparison, such as the influence of the socio-cultural and linguistic environments. It is necessary to add these characteristics of the voice because they are closely linked to the real-life context police services are faced with. For this purpose, a database (PTVVox, with two recordings conditions i.e micro and GSM conditions) recorded by the SDPTS has been created and will be exploited. Two types of complementary analyses will be conducted: a) an acoustic analysis where relevant cues will be extracted for voice characterization and b) perceptual experiments where the ability of listeners to discriminate voices will be tested. Acoustic analyses will contribute to delimit inter and intra speaker variability in order to test the robustness of these cues. Perceptual experiments will verify if the acoustic cues previously highlighted are actually used by listeners to identify speakers. These experiments will be conducted particularly in GSM condition (corresponding to the most common case in forensic cases). This project will allow the SDPTS labs to extend their ISO 17025 validation approaches to voice comparison. In order to disseminate the knowledge and the questions about voice comparison, seminars will be organized with the speech community and members of the judicial system. The project will also help to gather – and provide training for – a pool of young researchers specializing in voice comparison and the forensic applications thereof in order to compensate for a well-known lack of specialists in France.

In order to have a clearer understanding of the range of speech variations in both healthy and disordered speech populations, and in different speech conditions, the objective of this project is three-fold. It aims to draw a typological inventory of normal and abnormal speech variations, to test the adaptability of healthy and pathological speakers to different performance conditions (controlled to less controlled situations), and to test the conditioning effect of various linguistic-communicative constraints on these variations across populations. In parallel, our project aims to improve automatic speech processing systems confronting them to non prototypical speech (spontaneous or disordered speech). The examination of speech production in natural conditions has led to the observation of a large variety of non-prototypical pronunciations and has shown that these speech variations are not only the results of physiological or phonological factors but also of higher-level linguistic-constraints. Understanding the role of these constraints and the way they interact is challenging. Basic tenet of this proposal is that observation of disordered speech can inform the way normal speech is produced and vice-versa. Hence, one of the hypotheses raised in this project is that the forms of variation in disordered speech and their localization in the speech chain can inform the range and sites of possible normal variation. A comparison of how healthy and pathological speakers adapt their production to different speech conditions (here read vs. spontaneous speech) is also a way to better understand the processes and constraints at play during speech production. Two groups of patients suffering from motor speech disorders affecting different levels of production (planning in apraxia of speech and execution in dysarthia) are compared to healthy speakers. Read and spontaneous corpora will be designed and enriched at various levels ranging from phonetic alignment to discourse marker annotations. A combination of manual and automatic methods will be used to analyze the data. Outcomes of this research proposal will contribute to a better description of the form of variation occurring in the French language and to further explore the blurred boundary between normal and abnormal speech variations. It will improve our knowledge on both healthy and disordered speech variants, and on the speakers' ability to modulate their production to speech situations. This project is original in the way it proposes a new approach path to the question of the variability in the output of a healthy speech production system through a confrontation with the variability imposed/entailed by a disordered system. The two populations are not only considered as control conditions of each other but are studied in parallel. New contributions to the understanding of pathological signs in speech motor disorders and the definition of normal and abnormal speech patterns will be also achieved from the consideration of different speech styles, of the adaptability of speech performance, and on the role of the different constraints imposed to the production system in natural communication situations. Moreover, the proposed adaptation of approaches and methods developed for the study of healthy speech (elicitation and large corpora analysis techniques, multi-level annotations, automatic processing tools, etc.) to disordered speech constitutes a novel and stimulating methodological challenge. Finally, another interest of this project is its pluri-disciplinarity. The outstanding blend of competences in phonetics, phonology, prosody, clinical phonetics, clinical practice, speech engineering which are united in the proposed consortium are an essential part for the good progress of the project.

This proposed European research network will stand at the crossroads between language, social interaction, cognition, and the human brain. We believe that now is an appropriate moment for this network to arise, and that Europe offers a framework that will allow us to establish new collaborations to engage ourselves in very exciting scientific challenges. There is an extensive tradition of research on the relationships between language and the brain, but work in this domain has long been restricted to studying language production and comprehension in single individuals, using highly-controlled linguistic material. However, thanks to recent theoretical, methodological and empirical advances in language sciences, cognitive sciences and neurosciences, we are now in a position to extend our studies to ensembles of speakers engaged in conversational interactions. Our research now increasingly focuses on language in its primary site of occurrence, that is social interactions between people, and on how these interactions may both rely on and contribute to setting up brain-to-brain coupling relationships. We believe that a major turning point has been reached, in that social interactions between people are now the most appropriate framework for studying language and its cerebral and cognitive underpinnings.