Dans cet ouvrage, nous définissons précisément ce qu’est l’identification audio à grande échelle. En particulier, nous faisons une distinction entre l’identification exacte, destinée à rapprocher deux extraits sonores provenant d’un même enregistrement, et l’identification approchée, qui gère également la similarité musicale entre les signaux. A la lumière de ces définitions, nous concevons et examinons plusieurs modèles d’empreinte audio et évaluons leurs performances, tant en identification exacte qu’en identificationapprochée. / N this work we give a precise definition of large scale audio identification. In particular, we make a distinction between exact and approximate matching. In the first case, the goal is to match two signals coming from one same recording with different post-processings. In the second case, the goal is to match two signals that are musically similar. In light of these definitions, we conceive and evaluate different audio-fingerprint models.

Orientador: Cenira Maria Lupinacci / Resumo: Com o auxílio de diversos ramos da ciência, o ser humano buscou o desenvolvimento de suas demandas ao longo da história junto ao espaço natural, chegando para alguns autores, do ponto de vista do relevo, a ser denominado como um novo agente geomorfológico. A partir disso, as ações humanas vêm modificando o relevo em ambientes rurais, tanto para fins de agricultura como de pecuária. A fim de colaborar com os estudos que visam compreender os mecanismos de desenvolvimento de feições erosivas lineares e sua interação com agentes antropomórficos, teve-se como objetivo desta pesquisa apontar o potencial ao avanço erosivo linear de feições localizadas em uma bacia hidrográfica alocada no setor cuestiforme de Analândia (SP). Para o desenvolvimento da pesquisa, utilizaram-se as orientações de Mathias (2016), realizando o levantamento de diversas variáveis (granulometria, infiltração, resistência do solo à penetração, permeabilidade, declividade, fluxo acumulado de superfície e uso da terra) para formulação de cartas de síntese, as quais foram validadas a partir das áreas de ocorrência de feições erosivas e a partir dos dados de evolução do talude de uma voçoroca. Para validação, pautou-se na hipótese de que as classes com alto potencial ao avanço erosivo deveriam ter uma boa correspondência com os setores de maior concentração de feições erosivas e maior evolução do talude. Assim, tal sistemática permitiu identificar o setor central da bacia como uma zona de concentração de feições er... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: With help of several Science fields, human beings sought to develop their demands throughout history all over the natural space, reaching out to some authors, from the point of view of relief, to be named as a new geomorphological agent. From this, human actions have been modifying the relief in rural environments, for agriculture and livestock purposes. Seeking to collaborate with studies that aim to understand the mechanisms of development of linear erosive features and their interaction with anthropomorphic agents, this research goal was to indicate the potential to linear erosive rise of features located in a watershed located in the cuestiform sector from Analândia (SP). For the research development, it has been used the guidelines of Mathias (2016), performing the survey of several variables (particle size, infiltration, penetration resistance, permeability, slope, accumulated surface flow and land use) for the formulation of maps synthesis, which were validated from the areas of erosive features and from the evolution data of a gully erosion. For validation, it was based on the hypothesis that the kinds with high potential to erosive rise should have a good correspondence with sectors in higher concentration of erosive features and larger slope evolution. Thus, this systematic allowed to identify the central sector of the watershed as a zone of concentration of linear erosive features, which constitutes a transition area between the more resistant clay soils (to the no... (Complete abstract click electronic access below) / Mestre

No description available.

This MSc Thesis deals with face detection in image. In this approach, facial features (eyes, nose, mouth corners) are detected first and then joined to the whole face. For the facial features detection, classifiers trained with AdaBoost algorithm are used. Haar wavelets are used as features for classification.

In this thesis, we present the local rank differences (LRD). These novel image features are invariant to lighting changes and are suitable for object detection in programmable hardware, such as FPGA. The performance of AdaBoost classifiers with the LRD was tested on a face detection dataset with results which are similar to the Haar-like features which are the state of the art in real-time object detection. These results together with the fact that the LRD are evaluated much faster in FPGA then the Haar-like features are very encouraging and suggest that the LRD may be a solution for future hardware object detectors. We also present a framework for experiments with boosting methods in computer vision. This framework is very flexible and, at the same time, offers high learning performance and a possibility for future parallelization. The framework is available as open source software and we hope that it will simplify work for other researchers.

Presented thesis summarizes existing findings in the problematics of music recording comparison, in the field of so-called Music Information Retrieval. One of the goals of the thesis is to perform analysis of multiple musical renditions from the point of view of the tempo, dynamics and sound timbre. From the obtained data, such musical parameters will be chosen which exhibit the biggest potential to differentiate the origine of the individual renditions.

Die Studie zeigt die positiven affektiven Wirkungen des außerschulischen Lernortes Schülerlabor und repliziert damit frühere Arbeiten in diesem Bereich. Darüber hinaus wird der Einfluss von Betreuenden und Betreuung auf die Wirksamkeit von Schülerlaboren erstmals im Rahmen einer systematischen Untersuchung betrachtet und ein Wirkmodell abgeleitet. Die Besonderheit der vorliegenden Studie ist es, dass sie über diese allgemein formulierten Zusammenhänge hinaus aufzeigt, wie sich neben dem Professionswissen von Betreuenden auch die soziale Kompetenz auf die einzelnen Aspekte von wahrgenommener Betreuungsqualität auswirkt und wie diese wiederum affektive und kognitive Zielvariablen beeinflussen. Analysiert wurden die Angaben von 1490 Lernenden (Klassenstufe 9 bis 13 / Fragebogen im Pre-Post-Design) und 13 Betreuenden (Professionswissenstest). Methodisch erfolgte die Auswertung der Hypothesen durch Pre-Post-Vergleiche mittels Signifikanztests unter Angabe der Effektstärke. Die Untersuchung der Zusammenhänge wurde im Rahmen der Forschungsfrage durch Mehrebenenanalysen durchgeführt, mit denen Regressionen unter Beachtung der hierarchischen Struktur der Stichprobe möglich sind. Ein Strukturgleichungsmodell setzt die gefundenen Zusammenhänge miteinander in Verbindung und prüft diese im Sinne eines Gesamtmodells.:1. Einleitung Teil I: Theoretischer Rahmen & Forschungsstand 2. Lernort Schülerlabor 3. Forschungsgegenstand Teil II: Ziele & Konzeption der Studie 4. Hypothesen & Forschungsfrage 5. Konzeption & Methodik der Studie 6. Statistische Analyseverfahren 7. Prüfung der Erhebungsinstrumente Teil III: Empirische Forschungsergebnisse 8. Ergebnisse der Hauptuntersuchung 9. Zusammenfassung, Diskussion und Empfehlungen Anhang Verzeichnisse & Abspann

Spoken language is produced with a great deal of variability with which listeners must be able to cope. One source of variation is coarticulation, which is due to articulatory planning and transitions between segments. Recently, the temporal features of coarticulation were investigated during a picture/spoken-word matching task by using spliced stimuli carrying either congruent or incongruent subphonemic cues at the CV juncture (Archibald & Joanisse, 2011). ERPs were recorded with attention paid to the phonological mapping negativity (PMN) (Connolly & Phillips, 1994; Newman & Connolly, 2004) – a prelexical response sensitive to violations of phonological expectations. Results found that the PMN varied in response to coarticulation violations and concluded that phonetic features in spoken words influence prelexical processing during word recognition. Using a written-/spoken-word paradigm, Arbour, 2012 controlled phonological shape by using onsets that were either fricatives or stops, hypothesizing that coarticulatory information would be differentially processed due to their temporal differences. Findings supported the PMN’s sensitivity to coarticulation but also showed that temporal and physical differences between onsets modulated the effect. These results raise the question of whether acoustic distance between vowels will modulate prelexical processing of speech as reflected by the PMN amplitude: the focus of the current study. Words were organized into minimal sets such that all onset/coda combinations appeared with each vowel provided that English words resulted. Vowels were one of /i, u, æ, ɑ/, maximizing acoustic distance (height and backness). Data from 20 subjects indicate that the PMN is sensitive to the degree of difference between the original and post-splice vowels. When the number of distinctive features changing is greater, the result is an earlier, more robust PMN. This suggests that the rate of speech recognition is not static but dynamic, and is dependent on likeness of subphonemic features. / Thesis / Master of Science (MSc)

Fastighetsmarknaden har visat på ett växande behov av ett hjälpmedel för fastighetsvärderare som underlättar implementeringen av byggnaders gröna egenskaper i värderingsprocessen, samt för att investerare ska kunna se specifika gröna egenskapers effekt på fastighetsvärdet. Användningen av ett flertal olika miljöklassningssystem på marknaden skapar svårigheter vid jämförelse. RICS har tagit fram ett utkast på en hållbarhetschecklista för att komplettera värderingsutlåtandet med en utökad beskrivning av fastigheterna med en systematisk redovisning av egenskaper relevanta ur ett hållbarhetsperspektiv. Dess syfte är att tydliggöra skillnader och öka utbudet av tillgänglig transaktionsdata med en påföljande ökad transparens på fastighetsmarknaden. Studien syftar till att undersöka hur en hållbarhetschecklista kan tas fram, se ut och fungera, samt om den kan implementeras i ett projekt för en samhällsfastighet. Genom att använda mig av utkastet från RICS och intervjua aktörer på fastighetsmarknaden har en lista tagits fram som jämförts med miljöklassningssystemen och implementerats på ett specifikt projekt. Utfallet har sedan analyserats och legat till grund för de slutsatser som dragits. Resultatet av studien visar att den lista som utgår från RICS ”Sustainability Checklist” med relevanta hållbarhetsaspekter kan implementeras även på samhällsfastigheter då dessa innehar liknande egenskaper som de kommersiella fastigheterna, med vissa skillnader, som övergripande beror av: budgetrestriktioner, lokala förhållanden samt regleringar från kommunen och övriga parter. Även val av miljöklassningssystem har en viss påverkan på utfallet i fallstudien, då ett nära samband finns mellan parametrarna i hållbarhetschecklistan och miljöcertifieringssystemen, och där de kommersiella fastighetsutvecklarna i högre grad har valt att certifiera sina byggnader med de internationella systemen som är mer komplexa och innehar ett högre antal parametrar. Emellertid bygger en framgångsrik implementering av listan på att värderarens roll ändras till mer informerande, vilket också ställer krav på kunskapsnivå och utbudet av utbildning och vägledning för professionella värderare, en uppgift som ska upptas av professionella värderingsorgan. Slutligen fastställs att listan bör bearbetas ytterligare, med utveckling och förtydligande av vissa aspekter för att underlätta insamling av information samt underlätta för värderare att i praktiken kunna använda den som ett komplement i värderingsarbetet. / The real estate market has shown an increased demand among real estate appraisers for a tool to facilitate the implementation of green building features into the valuation process, and for investors to discern a specific green feature’s impact upon value. The wide range of green building rating systems used on the real estate market today causes difficulties in comparison. To supplement the valuation report with an extended description of the properties including a systematic presentation of relevant features from a sustainability perspective, RICS (2012) has formed a draft of a ”sustainability checklist”, which purpose is to highlight differences, increase the supply of information and enhance transparency in the property market. The study aims to investigate the development, design and function of a sustainability checklist, and whether it can be implemented on public properties. A list was created based on the draft from RICS ”Sustainability Checklist”, and interviewed players in the real estate market, which were implemented on a specific project consisting of a public building. The result of the study was analysed and has formed the basis of the conclusions drawn. The result of the study conducted within this thesis shows that the list based on RICS ”Sustainability Checklist” with sustainability aspects included can be implemented also on public buildings, as these buildings possess similar features as the commercial buildings. Some main differences can be seen, overall resulting from: budget constraints, local conditions and regulations from the municipality and other parties. Also, the choice of a certain green building rating system influence the effect on the outcome, whereas a close relationship can be find between certain aspects within the sustainability checklist and green building rating systems, and the commercial real estate developers increasingly chose to certify their buildings with the international systems, which are more complex and contain a higher number of aspects. However, in order to achieve a successful implementation of the list among appraisers within the real estate market, the role of the valuer must change to more informative, which also make demands on the level of knowledge and the provision of training and guidance for professional appraisers, a task to be undertaken by professional valuators. Furthermore, it is also established within this study that the list should be further processed and modified in order to facilitate the gathering of information and to simplify for the appraiser to use it as a supplement within the valuation process.

While much research has shown that some linguistic features can indicate a person is lying, this line of research has led to conflicting results. Furthermore, very little research has been done to verify that these supposed linguistic features of deception are universal. In addition, few studies have researched the cross-cultural perceptions of deception, which knowledge could greatly improve the detection of deception across cultures. The current study addresses these gaps in the literature by analyzing and comparing truthful and deceptive discourse of eight native English-speaking Americans and eight non-native English-speaking Ghanaians. The discourse was elicited in one-on-one interviews where each interviewee spontaneously responded to questions about themselves. Later, interviewee responses were judged by 47 native English-speaking Americans and 35 non-native English-speaking Ghanaians. The results showed that Americans and Ghanaians lie differently—Americans' lies were more superfluous and redundant; had more pronoun inconsistencies, adjectives, adverbs, and modal verbs; and had fewer negative emotion words than their truths. Ghanaians' lies, on the other hand, also had more pronoun inconsistencies but had fewer negations than their truths. Furthermore, the groups' baseline speech differed in superfluousness, positive emotion words, word count, and response latency. Regarding detecting deception, Ghanaians were slightly more accurate and significantly more confident in detecting lies than Americans. Both groups were slightly more accurate and confident in judging the veracity of statements within their own cultures. Neither group, however, demonstrated truth- or lie-bias cross-culturally. These results have implications for law enforcement investigators and analysts who can learn the differences between Americans' and Ghanaians' truthful and deceptive speech so as to more accurately detect deception through language. In addition, cross-cultural deception perception research can improve cross-cultural communication and understanding.