<p>In this paper, a navigation support approach for speech-only interaction based on auditory icons for room-based designs is presented, i.e. naturally occurring sounds that have a natural mapping to the system's underlying design metaphor. In contrast to many recent investigations that have focused on multi-modal or augmented reality systems, this paper concerns a unimodal speech and sound-only system. An auditory icon based prototype system for buildings maintenance support using a room-based metaphor was developed. The design was evaluated in a comparison with earcons and no-sound designs. The users’ subjective attitudes toward auditory icons were significantly more positive than to earcons.</p>

In this paper, a navigation support approach for speech-only interaction based on auditory icons for room-based designs is presented, i.e. naturally occurring sounds that have a natural mapping to the system's underlying design metaphor. In contrast to many recent investigations that have focused on multi-modal or augmented reality systems, this paper concerns a unimodal speech and sound-only system. An auditory icon based prototype system for buildings maintenance support using a room-based metaphor was developed. The design was evaluated in a comparison with earcons and no-sound designs. The users’ subjective attitudes toward auditory icons were significantly more positive than to earcons.

Modern teknik som intelligenta personliga assistenter (IPA) som finns i smartphones kan förenkla människors vardagliga liv. De kan hjälpa personer med funktionshinder och människor som lever stressfulla liv. Genom tal, kan IPA:er enkelt hjälpa personer som ofta har händerna fulla att hitta vägbeskrivningar, information eller skicka meddelanden över hela världen. För att locka människor till att förlita sig på IPA:er i det dagliga livet räcker det inte att den är intelligent och tilltalande, den måste också vara tillförlitlig. Det finns många faktorer som är involverade i att bygga ett trovärdigt system. Denna uppsats fokuserar på att ta reda på vilka egenskaper som gör en IPA:s röst trovärdig och vilken typ av röster designers bör göra tillgängliga för sina användare. Kvalitativa och kvantitativa forskningsmetoder har använts i studien. Sammanlagt 11 semi-strukturerade intervjuer genomfördes och varje innehöll ett kort användartest. 100 personer deltog i en enkätundersökning med ljudklipp och specifika frågor om egenskaper hos pålitliga IPA röster. 70 av 100 tillfrågade var överens om att röster som medför känslor, speciellt en känsla av trygghet och lugn, gör en IPA-röst trovärdig. Att låta tydlig, behaglig, människoliknande och erfaren, samt att ha en dialekt som användaren kan relatera till anses också vara viktiga röstegenskaper. Kön anses vara den minst viktiga. / Modern technologies such as intelligent personal assistants (IPA) found in smartphones can simplify the daily lives of people. They are not only useful for people with disabilities but also for those who live stressful lives. Through speech, IPAs can easily help a busy full-handed person find directions, information or send messages across the world. To entice people into relying on IPAs in their daily life, it not only has to be intelligent or appealing enough but also trustworthy. There are many principles involved in building a trustworthy system. This paper focuses on finding out what characteristics make an IPA-voice trustworthy and what kind of voices designers should make available for their users. Qualitative and quantitative research methods were used in the study. A total of 11 semi-structured interviews were carried out each in conjunction with a brief user test. 100 people participated in a survey with audio clips and specific questions regarding characteristics of trustworthy IPA-voices. 70 of 100 respondents agreed that voices that bring about emotions, particularly the feeling of security and tranquility, make an IPA-voice trustworthy. To sound clear, pleasant, human-like and experienced, and have a dialect that the user can relate to are also considered important voice qualities. Gender is considered the least important.

Our lives become longer with new advancements in medicine and technology. As a result, cognition and memory impairments will become a growing concern[23]. To be able to treat cognitive decline, dementia and related diseases, the need for early detection of memory deficiencies increases. This study aimed to validate the use of a new digital interaction version of a common memory test, the Rey Auditory Verbal Learning Test (RAVLT), compared with norm from the traditional human-to-human interaction version of the same test. This was performed through userstudies with both test versions in a counterbalanced order. The results consisted of quantitative results from the tests together with qualitative data from interviews with the participants in connection to each test occasion. The results showed a slight difference in some aspects of the quantitative test results, but at the same time, no significant difference for most of the measured results. Further, the interviews demonstrated that the differences in interaction and experience between the versions of the tests could be the reason for some of the differences in the quantitative result. The conclusion from this study was that there are many difficult problems to solve for speech interfaces in neurocognitive tests like the RAVLT. In line with previous research [22], the results highlight the importance of a natural speech interaction that conforms with the basic principles of human conversation, to create a stress-free experience and test results that can be reliable and comparable.

Esta dissertação tem por objetivo descrever e analisar construções de tópico no português falado no município do Libolo/Angola (doravante, PLB). Tais construções de tópico podem ser caracterizadas, pragmaticamente, por certo sintagma/sentença que está ligado a uma sentença-comentário, sem, contudo, carregar a força ilocucionária do comentário; sintaticamente, o tópico pode ser caracterizado como um tipo de codificação de informação ligada ao sintagma complementizador (CP). Esta pesquisa, que tem como ponto de partida trabalhos sobre o tópico no português brasileiro (PB) e no português europeu (PE), apresenta um primeiro estudo do tópico em PLB, apontando, ainda, uma proposta inicial de aspectos da estrutura gramatical nesta variedade de português falada na África. A pesquisa lança, portanto, novas luzes em estudos inseridos em um projeto maior, denominado Projeto Libolo ver, entre outros, Figueiredo e Oliveira (2016) , cujo corpus centra-se em dados coletados em Angola nos anos de 2011, 2013 e 2016. / This dissertation consists of a study on topic constructions in the Portuguese spoken in the municipality of Libolo/Angola (hereinafter PLB). Such topic constructions can be characterized, pragmatically, by a certain phrase/sentence that is connected to a comment sentence, without, however, carrying the illocutionary force of the comment; syntactically, the topic can be characterized as a type of information coding linked to the Complementizer Phrase (CP). The results of the research here presented, which has as its starting point the studies on topic constructions in Brazilian Portuguese (PB) and European Portuguese (PE), present a first descriptive/explanatory proposal of the topic constructions in PLB, and points out also an initial proposal of structural aspects of this variety of Portuguese spoken in Africa. The research, therefore, throws new light on studies inserted in a larger project, called \"Project Libolo\" see, among others, Figueiredo and Oliveira (2016) whose corpus focuses on data collected in Angola in 2011, 2013 and 2016.

Esta dissertação tem por objetivo descrever e analisar construções de tópico no português falado no município do Libolo/Angola (doravante, PLB). Tais construções de tópico podem ser caracterizadas, pragmaticamente, por certo sintagma/sentença que está ligado a uma sentença-comentário, sem, contudo, carregar a força ilocucionária do comentário; sintaticamente, o tópico pode ser caracterizado como um tipo de codificação de informação ligada ao sintagma complementizador (CP). Esta pesquisa, que tem como ponto de partida trabalhos sobre o tópico no português brasileiro (PB) e no português europeu (PE), apresenta um primeiro estudo do tópico em PLB, apontando, ainda, uma proposta inicial de aspectos da estrutura gramatical nesta variedade de português falada na África. A pesquisa lança, portanto, novas luzes em estudos inseridos em um projeto maior, denominado Projeto Libolo ver, entre outros, Figueiredo e Oliveira (2016) , cujo corpus centra-se em dados coletados em Angola nos anos de 2011, 2013 e 2016. / This dissertation consists of a study on topic constructions in the Portuguese spoken in the municipality of Libolo/Angola (hereinafter PLB). Such topic constructions can be characterized, pragmatically, by a certain phrase/sentence that is connected to a comment sentence, without, however, carrying the illocutionary force of the comment; syntactically, the topic can be characterized as a type of information coding linked to the Complementizer Phrase (CP). The results of the research here presented, which has as its starting point the studies on topic constructions in Brazilian Portuguese (PB) and European Portuguese (PE), present a first descriptive/explanatory proposal of the topic constructions in PLB, and points out also an initial proposal of structural aspects of this variety of Portuguese spoken in Africa. The research, therefore, throws new light on studies inserted in a larger project, called \"Project Libolo\" see, among others, Figueiredo and Oliveira (2016) whose corpus focuses on data collected in Angola in 2011, 2013 and 2016.

Η παρούσα διδακτορική διατριβή πραγματεύεται προβλήματα που αφορούν το χώρο της τεχνολογίας ομιλίας, με στόχο τη αναγνώριση συναισθημάτων από ομιλία με χρήση τεχνικών ψηφιακής επεξεργασίας σήματος και μηχανικής μάθησης. Πιο αναλυτικά, στα πλαίσια της διατριβής προτάθηκαν και μελετήθηκαν καινοτόμες μέθοδοι σε μια σειρά από εφαρμογές που αξιοποιούν σύστημα αναγνώρισης συναισθηματικών καταστάσεων από ομιλία. Ο βασικός στόχος των μεθόδων ήταν η αντιμετώπιση των προκλήσεων που παρουσιάζονται όταν ένα σύστημα αναγνώρισης συναισθηματικών καταστάσεων καλείται να λειτουργήσει σε πραγματικές συνθήκες, με αυθόρμητες αντιδράσεις, ανεξαρτήτως ομιλητή. Πιο συγκεκριμένα, στα πλαίσια της διατριβής, αξιολογήθηκε η συμπεριφορά ενός συστήματος αναγνώρισης συναισθημάτων σε προσποιητή ομιλία και σε διαφορετικές συνθήκες θορύβου, και συγκρίθηκε η απόδοση του συστήματος με την υποκειμενική αξιολόγηση των ακροατών. Επιπλέον, περιγράφηκε ο σχεδιασμός και η υλοποίηση βάση δεδομένων συναισθηματικής ομιλίας, όπως αυτή προκύπτει από την αλληλεπίδραση μη-έμπειρων χρηστών με ένα διαλογικό σύστημα και προτάθηκε ένα σύστημα το οποίο εντοπίζει αρνητικές συναισθηματικές καταστάσεις, στο ανεξάρτητου ομιλητή πρόβλημα, με χρήση μοντέλου Γκαουσιανών κατανομών. Η προτεινόμενη αρχιτεκτονική συνδυάζει παραμέτρους ομιλίας χαμηλού και υψηλού επιπέδου και εφαρμόζεται στα πραγματικά δεδομένα. Επίσης, αξιολογήθηκε και υλοποιήθηκε η πρακτική εφαρμογή ενός συστήματος αναγνώρισης συναισθημάτων βασισμένου σε οικουμενικό μοντέλο Γκαουσιανών κατανομών σε διαφορετικούς τύπους δεδομένων πραγματικής ζωής. Ακόμα, παρουσιάστηκε μια πρωτότυπη αρχιτεκτονική κατηγοριοποίησης για αναγνώριση συνυπαρχόντων συναισθημάτων από ομιλία προερχόμενη από αλληλεπίδραση σε πραγματικά περιβάλλοντα. Σε αντίθεση με γνωστές προσεγγίσεις, η προτεινόμενη αρχιτεκτονική μοντελοποιεί τις συνυπάρχουσες συναισθηματικές καταστάσεις μέσω της κατασκευής μιας πολυσταδιακής αρχιτεκτονικής κατηγοριοποίησης. Τα πειραματικά αποτελέσματα που διενεργήθηκαν υποδεικνύουν ότι η προτεινόμενη αρχιτεκτονική είναι πλεονεκτική για τις συναισθηματικές καταστάσεις που είναι πιο διαχωρίσιμες, γεγονός που οδηγεί σε βελτίωση της συνολικής απόδοσης του συστήματος. / In this doctoral dissertation a number of novel approaches were proposed and evaluated in different applications that utilize emotion awareness. The major target of the proposed methods was facing the difficulties existing, when an emotion recognition system is asked to operate in real-life conditions, where human speech is characterized by spontaneous and genuine formulations. In detail, within the present dissertation, the performance of an emotion recognition system was evaluated, initially, in acted speech, under different noise conditions, and this performance was compared to the one of human listeners. Further, the design and implementation of a real world emotional speech corpus is described, as this results from the interaction of naive users with a smart home dialogue system. Moreover, a system which utilizes low and high level descriptors was suggested. The suggested architecture leads to significantly better performance in some working points of the integrated system in the dialogue system. Furthermore, we propose a novel multistage classification scheme for affect recognition from real-life speech. In contrast with conventional approaches for affect/emotion recognition from speech, the proposed scheme models co-occurring affective states by constructing a multistage classification scheme. The empirical experiments performed indicate that the proposed classification scheme offers an advantage for those classes that are more separable, which contributes for improving the overall performance of the affect recognition system.