Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

18.08 German language. --- Deutschunterricht. --- German language --- German. --- Minimalpaar. --- Phonemanalyse. --- Phonemes. --- Phonetik. --- Phonologie. --- Phonemics. --- Deutsch.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

La parole de synthèse peut-être réalisée via diverses méthodes, notamment la synthèse articulatoire. Il existe différentes modélisations physiques : le modèle acoustique unidimensionnel (1D), le modèle acoustique tridimensionnel (3D) et le modèle d’algorithme d’extension (BWE). Le modèle 3D semble offrir la parole la plus naturelle (Gully, 2017). D’une part, il se base sur la forme précise du tractus vocal, générant des simulations acoustiques proches de la réalité, et d’autre part, il permet de modéliser correctement les hautes fréquences (HF) (> 5kHz) (Arnela et al., 2019 ; Freixes et al., 2018). Longtemps mises de côté dans les recherches sur la perception de la parole, ces HF connaissent un nouvel intérêt depuis plusieurs années, car elles semblent jouer un rôle important dans l’aspect naturel de la parole (Vitela et al., 2015 ; Monson & Caravello, 2019 ; Boyd-Pratt & Donnai, 2020 ; Birkholz & Drechsel, 2021). Ce mémoire s’inscrit dans un projet de développement d’un outil de synthèse articulatoire à large bande, dont l’aspect se veut le plus naturel possible. Notre objectif est de déterminer, pour la synthèse articulatoire, comment les différents modèles : 1D, 3D et BWE, impactent la perception de l’aspect naturel de la parole chez les jeunes adultes. Après avoir rempli un questionnaire anamnestique et passé une audiométrie tonale, 40 participants ont réalisé deux tâches expérimentales. La première tâche était une comparaison par paires, qui avait pour but de comparer l’aspect naturel des différents stimuli deux à deux. La seconde tâche était une évaluation de l’aspect naturel des stimuli, à l’aide d’une échelle métrique allant de 0 (pas du tout naturel) à 100 (totalement naturel). Ces tâches nous ont permis de répondre à trois hypothèses concernant le degré de réalisme physique des modèles acoustiques, et d’investiguer différentes questions de recherche concernant le genre de la voix de synthèse, la qualité vocale, et le type de phonème, et la fiabilité inter-juges. Les deux tâches expérimentales ont permis de mettre en avant plusieurs effets significatifs. Un effet significatif du modèle acoustique a été trouvé, de façon générale, le modèle 3D est plus naturel. Un effet significatif du type de phonème a montré que le degré de naturalité dépend du phonème. Une interaction a été trouvée entre le modèle acoustique et le type de phonème, révélant que l’aspect naturel des modèles diffère selon le type de phonème. Seule la seconde tâche expérimentale a permis de mettre en lumière un effet significatif du genre de la voix de synthèse, indiquant que la voix de synthèse masculine parait plus naturelle que la féminine. Ce mémoire a cherché à explorer le rôle des HF dans la perception des phonèmes selon différents degrés de réalismes physiques de modèles acoustiques.

Hautes fréquences --- Perception --- Phonèmes --- Synthèse de la parole --- Modélisations physiques --- Modèles acoustiques --- Sciences sociales & comportementales, psychologie > Multidisciplinaire, généralités & autres

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Fonemen. Taalanalyse. (Congres) --- Spraak. Spectroanalyse. (Congres) --- Fonetiek (Experimentele). (Congres) --- Son. Enregistrement /et phonétique. (Congrès) --- Phonèmes. Analyse linguistique. (Congrès) --- Parole. Spectrographie acoustique. (Congrès) --- Phonétique expérimentale. (Congrès) --- Geluid. Opname /en fonetiek. (Congres)

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

One of the most promising developments in modelling knowledge is cognitive network science, which aims to investigate cognitive phenomena driven by the networked, associative organization of knowledge. For example, investigating the structure of semantic memory via semantic networks has illuminated how memory recall patterns influence phenomena such as creativity, memory search, learning, and more generally, knowledge acquisition, exploration, and exploitation. In parallel, neural network models for artificial intelligence (AI) are also becoming more widespread as inferential models for understanding which features drive language-related phenomena such as meaning reconstruction, stance detection, and emotional profiling. Whereas cognitive networks map explicitly which entities engage in associative relationships, neural networks perform an implicit mapping of correlations in cognitive data as weights, obtained after training over labelled data and whose interpretation is not immediately evident to the experimenter. This book aims to bring together quantitative, innovative research that focuses on modelling knowledge through cognitive and neural networks to gain insight into mechanisms driving cognitive processes related to knowledge structuring, exploration, and learning. The book comprises a variety of publication types, including reviews and theoretical papers, empirical research, computational modelling, and big data analysis. All papers here share a commonality: they demonstrate how the application of network science and AI can extend and broaden cognitive science in ways that traditional approaches cannot.

Information technology industries --- text mining --- big data --- analytics --- review --- self-organization --- computational philosophy --- brain --- synaptic learning --- adaptation --- functional plasticity --- activity-dependent resonance states --- circular causality --- somatosensory representation --- prehensile synergies --- robotics --- COVID-19 --- social media --- hashtag networks --- emotional profiling --- cognitive science --- network science --- sentiment analysis --- computational social science --- Twitter --- VADER scoring --- correlation --- semantic network analysis --- intellectual disability --- adolescents --- EEG --- emotional states --- working memory --- depression --- anxiety --- graph theory --- classification --- machine learning --- neural networks --- phonotactic probability --- neighborhood density --- sub-lexical representations --- lexical representations --- phonemes --- biphones --- cognitive network --- smart assistants --- knowledge generation --- intelligent systems --- web components --- deep learning --- web-based interaction --- cognitive network science --- text analysis --- natural language processing --- artificial intelligence --- emotional recall --- cognitive data --- AI --- pharmacological text corpus --- automatic relation extraction --- gender stereotypes --- story tropes --- movie plots --- network analysis --- word co-occurrence network --- n/a

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

One of the most promising developments in modelling knowledge is cognitive network science, which aims to investigate cognitive phenomena driven by the networked, associative organization of knowledge. For example, investigating the structure of semantic memory via semantic networks has illuminated how memory recall patterns influence phenomena such as creativity, memory search, learning, and more generally, knowledge acquisition, exploration, and exploitation. In parallel, neural network models for artificial intelligence (AI) are also becoming more widespread as inferential models for understanding which features drive language-related phenomena such as meaning reconstruction, stance detection, and emotional profiling. Whereas cognitive networks map explicitly which entities engage in associative relationships, neural networks perform an implicit mapping of correlations in cognitive data as weights, obtained after training over labelled data and whose interpretation is not immediately evident to the experimenter. This book aims to bring together quantitative, innovative research that focuses on modelling knowledge through cognitive and neural networks to gain insight into mechanisms driving cognitive processes related to knowledge structuring, exploration, and learning. The book comprises a variety of publication types, including reviews and theoretical papers, empirical research, computational modelling, and big data analysis. All papers here share a commonality: they demonstrate how the application of network science and AI can extend and broaden cognitive science in ways that traditional approaches cannot.

Information technology industries --- text mining --- big data --- analytics --- review --- self-organization --- computational philosophy --- brain --- synaptic learning --- adaptation --- functional plasticity --- activity-dependent resonance states --- circular causality --- somatosensory representation --- prehensile synergies --- robotics --- COVID-19 --- social media --- hashtag networks --- emotional profiling --- cognitive science --- network science --- sentiment analysis --- computational social science --- Twitter --- VADER scoring --- correlation --- semantic network analysis --- intellectual disability --- adolescents --- EEG --- emotional states --- working memory --- depression --- anxiety --- graph theory --- classification --- machine learning --- neural networks --- phonotactic probability --- neighborhood density --- sub-lexical representations --- lexical representations --- phonemes --- biphones --- cognitive network --- smart assistants --- knowledge generation --- intelligent systems --- web components --- deep learning --- web-based interaction --- cognitive network science --- text analysis --- natural language processing --- artificial intelligence --- emotional recall --- cognitive data --- AI --- pharmacological text corpus --- automatic relation extraction --- gender stereotypes --- story tropes --- movie plots --- network analysis --- word co-occurrence network --- text mining --- big data --- analytics --- review --- self-organization --- computational philosophy --- brain --- synaptic learning --- adaptation --- functional plasticity --- activity-dependent resonance states --- circular causality --- somatosensory representation --- prehensile synergies --- robotics --- COVID-19 --- social media --- hashtag networks --- emotional profiling --- cognitive science --- network science --- sentiment analysis --- computational social science --- Twitter --- VADER scoring --- correlation --- semantic network analysis --- intellectual disability --- adolescents --- EEG --- emotional states --- working memory --- depression --- anxiety --- graph theory --- classification --- machine learning --- neural networks --- phonotactic probability --- neighborhood density --- sub-lexical representations --- lexical representations --- phonemes --- biphones --- cognitive network --- smart assistants --- knowledge generation --- intelligent systems --- web components --- deep learning --- web-based interaction --- cognitive network science --- text analysis --- natural language processing --- artificial intelligence --- emotional recall --- cognitive data --- AI --- pharmacological text corpus --- automatic relation extraction --- gender stereotypes --- story tropes --- movie plots --- network analysis --- word co-occurrence network