1,721,109 research outputs found
Negotiation Models for Mobile Tactile Interaction
No full textWith the recent introduction of mass-market mobile phones with touch-sensitive displays, location, bearing and motion sensing, we are on the cusp of significant progress in a highly interactive mobile social networking. We propose that such systems must work in various contexts, levels of uncertainties and utilize different types of human senses. In order to explore the feasibility of such a system we describe an experiment with a multimodal implementation which allows users to engage in a continuous interaction with each other by using capacitive touch input, visual and/or vibro-tactile feedback and perform a goal-oriented collaborative task of target acquisition. Initial user study found the approach to be interesting and engaging despite the constraints imposed by the interaction method
Speaker diarization of multi-party conversations using participants role information: political debates and professional meetings
Full text linkSpeaker Diarization aims at inferring who spoke when in an audio stream and involves two simultaneous unsupervised tasks: (1) the estimation of the number of speakers, and (2) the association of speech segments to each speaker. Most of the recent efforts in the domain have addressed the problem using machine learning techniques or statistical methods (for a review see [11]) ignoring the fact that the data consists of instances of human conversations
Mobile phones and social signal processing for analysis and understanding of dyadic conversations
Full text linkSocial Signal Processing is the domain aimed at bridging the social intelligence gap between humans and machines via modeling, analysis and synthesis of nonverbal behavior in social interactions. One of the main challenges of the domain is to sense unobtrusively the behavior of social interaction participants, one of the key conditions to preserve the spontaneity and naturalness of the interactions under exam. In this respect, mobile devices offer a major opportunity because they are equipped with a wide array of sensors that, while capturing the behavior of their users with an unprecedented depth, are still invisible. This is particularly important because mobile devices are part of the everyday life of a large number of individuals and, hence, they can be used to investigate and sense natural and spontaneous scenarios
Haptic feedback: a brief history from telepresence to virtual reality
No full textThis paper presents a short review of the history surrounding the development of haptic feedback systems, from early manipulators and telerobots, used in the nuclear and subsea industries, to today’s impressive desktop devices, used to support real-time interaction with 3D visual simulations, or Virtual Reality. Four examples of recent VR projects are described, illustrating the use of haptic feedback in ceramics, aerospace, surgical and defence applications. These examples serve to illustrate the premise that haptic feedback systems have evolved much faster than their visual display counterparts and are, today, delivering impressive peripheral devices that are truly usable by non-specialist users of computing technology.</p
Erottelevien mallien oppiminen rakenteellisesta monianturitiedosta ihmiseen liittyvän tilannetiedon tunnistuksessa
No full textAbstractIn this work, statistical machine learning and pattern recognition methods were developed and applied to sensor-based human context recognition. More precisely, we concentrated on an effective discriminative learning framework, where input-output mapping is learned directly from a labeled dataset. Non-parametric discriminative classification and regression models based on kernel methods were applied. They include support vector machines (SVM) and Gaussian processes (GP), which play a central role in modern statistical machine learning. Based on these established models, we propose various extensions for handling structured data that usually arise from real-life applications, for example, in a field of context-aware computing. We applied both SVM and GP techniques to handle data with multiple classes in a structured multi-sensor domain. Moreover, a framework for combining data from several sources in this setting was developed using multiple classifiers and fusion rules, where kernel methods are used as base classifiers. We developed two novel methods for handling sequential input and output data. For sequential time-series data, a novel kernel based on graphical presentation, called a weighted walk-based graph kernel (WWGK), is introduced. For sequential output labels, discriminative temporal smoothing (DTS) is proposed. Again, the proposed algorithms are modular, so different kernel classifiers can be used as base models. Finally, we propose a group of techniques based on Gaussian process regression (GPR) and particle filtering (PF) to learn to track multiple targets. We applied the proposed methodology to three different human-motion-based context recognition applications: person identification, person tracking, and activity recognition, where floor (pressure-sensitive and binary switch) and wearable acceleration sensors are used to measure human motion and gait during walking and other activities. Furthermore, we extracted a useful set of specific high-level features from raw sensor measurements based on time, frequency, and spatial domains for each application. As a result, we developed practical extensions to kernel-based discriminative learning to handle many kinds of structured data applied to human context recognition. TiivistelmäTässä työssä kehitettiin ja sovellettiin tilastollisen koneoppimisen ja hahmontunnistuksen menetelmiä anturipohjaiseen ihmiseen liittyvän tilannetiedon tunnistamiseen. Esitetyt menetelmät kuuluvat erottelevan oppimisen viitekehykseen, jossa ennustemalli sisääntulomuuttujien ja vastemuuttujan välille voidaan oppia suoraan tunnetuilla vastemuuttujilla nimetystä aineistosta. Parametrittomien erottelevien mallien oppimiseen käytettiin ydinmenetelmiä kuten tukivektorikoneita (SVM) ja Gaussin prosesseja (GP), joita voidaan pitää yhtenä modernin tilastollisen koneoppimisen tärkeimmistä menetelmistä. Työssä kehitettiin näihin menetelmiin liittyviä laajennuksia, joiden avulla rakenteellista aineistoa voidaan mallittaa paremmin reaalimaailman sovelluksissa, esimerkiksi tilannetietoisen laskennan sovellusalueella. Tutkimuksessa sovellettiin SVM- ja GP-menetelmiä moniluokkaisiin luokitteluongelmiin rakenteellisen monianturitiedon mallituksessa. Useiden tietolähteiden käsittelyyn esitetään menettely, joka yhdistää useat opetetut luokittelijat päätöstason säännöillä lopulliseksi malliksi. Tämän lisäksi aikasarjatiedon käsittelyyn kehitettiin uusi graafiesitykseen perustuva ydinfunktio sekä menettely sekventiaalisten luokkavastemuuttujien käsittelyyn. Nämä voidaan liittää modulaarisesti ydinmenetelmiin perustuviin erotteleviin luokittelijoihin. Lopuksi esitetään tekniikoita usean liikkuvan kohteen seuraamiseen. Menetelmät perustuvat anturitiedosta oppivaan GP-regressiomalliin ja partikkelisuodattimeen. Työssä esitettyjä menetelmiä sovellettiin kolmessa ihmisen liikkeisiin liittyvässä tilannetiedon tunnistussovelluksessa: henkilön biometrinen tunnistaminen, henkilöiden seuraaminen sekä aktiviteettien tunnistaminen. Näissä sovelluksissa henkilön asentoa, liikkeitä ja astuntaa kävelyn ja muiden aktiviteettien aikana mitattiin kahdella erilaisella paineherkällä lattia-anturilla sekä puettavilla kiihtyvyysantureilla. Tunnistusmenetelmien laajennuksien lisäksi jokaisessa sovelluksessa kehitettiin menetelmiä signaalin segmentointiin ja kuvaavien piirteiden irroittamiseen matalantason anturitiedosta. Tutkimuksen tuloksena saatiin parannuksia erottelevien mallien oppimiseen rakenteellisesta anturitiedosta sekä erityisesti uusia menettelyjä tilannetiedon tunnistamiseen. Academic dissertation to be presented with the assent of the Doctoral Training Committee of Technology and Natural Sciences of the University of Oulu for public defence in Auditorium TS101, Linnanmaa, on 27 June 2012, at 12 noonAbstract
In this work, statistical machine learning and pattern recognition methods were developed and applied to sensor-based human context recognition. More precisely, we concentrated on an effective discriminative learning framework, where input-output mapping is learned directly from a labeled dataset. Non-parametric discriminative classification and regression models based on kernel methods were applied. They include support vector machines (SVM) and Gaussian processes (GP), which play a central role in modern statistical machine learning. Based on these established models, we propose various extensions for handling structured data that usually arise from real-life applications, for example, in a field of context-aware computing.
We applied both SVM and GP techniques to handle data with multiple classes in a structured multi-sensor domain. Moreover, a framework for combining data from several sources in this setting was developed using multiple classifiers and fusion rules, where kernel methods are used as base classifiers. We developed two novel methods for handling sequential input and output data. For sequential time-series data, a novel kernel based on graphical presentation, called a weighted walk-based graph kernel (WWGK), is introduced. For sequential output labels, discriminative temporal smoothing (DTS) is proposed. Again, the proposed algorithms are modular, so different kernel classifiers can be used as base models. Finally, we propose a group of techniques based on Gaussian process regression (GPR) and particle filtering (PF) to learn to track multiple targets.
We applied the proposed methodology to three different human-motion-based context recognition applications: person identification, person tracking, and activity recognition, where floor (pressure-sensitive and binary switch) and wearable acceleration sensors are used to measure human motion and gait during walking and other activities. Furthermore, we extracted a useful set of specific high-level features from raw sensor measurements based on time, frequency, and spatial domains for each application. As a result, we developed practical extensions to kernel-based discriminative learning to handle many kinds of structured data applied to human context recognition.Tiivistelmä
Tässä työssä kehitettiin ja sovellettiin tilastollisen koneoppimisen ja hahmontunnistuksen menetelmiä anturipohjaiseen ihmiseen liittyvän tilannetiedon tunnistamiseen. Esitetyt menetelmät kuuluvat erottelevan oppimisen viitekehykseen, jossa ennustemalli sisääntulomuuttujien ja vastemuuttujan välille voidaan oppia suoraan tunnetuilla vastemuuttujilla nimetystä aineistosta. Parametrittomien erottelevien mallien oppimiseen käytettiin ydinmenetelmiä kuten tukivektorikoneita (SVM) ja Gaussin prosesseja (GP), joita voidaan pitää yhtenä modernin tilastollisen koneoppimisen tärkeimmistä menetelmistä. Työssä kehitettiin näihin menetelmiin liittyviä laajennuksia, joiden avulla rakenteellista aineistoa voidaan mallittaa paremmin reaalimaailman sovelluksissa, esimerkiksi tilannetietoisen laskennan sovellusalueella.
Tutkimuksessa sovellettiin SVM- ja GP-menetelmiä moniluokkaisiin luokitteluongelmiin rakenteellisen monianturitiedon mallituksessa. Useiden tietolähteiden käsittelyyn esitetään menettely, joka yhdistää useat opetetut luokittelijat päätöstason säännöillä lopulliseksi malliksi. Tämän lisäksi aikasarjatiedon käsittelyyn kehitettiin uusi graafiesitykseen perustuva ydinfunktio sekä menettely sekventiaalisten luokkavastemuuttujien käsittelyyn. Nämä voidaan liittää modulaarisesti ydinmenetelmiin perustuviin erotteleviin luokittelijoihin. Lopuksi esitetään tekniikoita usean liikkuvan kohteen seuraamiseen. Menetelmät perustuvat anturitiedosta oppivaan GP-regressiomalliin ja partikkelisuodattimeen.
Työssä esitettyjä menetelmiä sovellettiin kolmessa ihmisen liikkeisiin liittyvässä tilannetiedon tunnistussovelluksessa: henkilön biometrinen tunnistaminen, henkilöiden seuraaminen sekä aktiviteettien tunnistaminen. Näissä sovelluksissa henkilön asentoa, liikkeitä ja astuntaa kävelyn ja muiden aktiviteettien aikana mitattiin kahdella erilaisella paineherkällä lattia-anturilla sekä puettavilla kiihtyvyysantureilla. Tunnistusmenetelmien laajennuksien lisäksi jokaisessa sovelluksessa kehitettiin menetelmiä signaalin segmentointiin ja kuvaavien piirteiden irroittamiseen matalantason anturitiedosta. Tutkimuksen tuloksena saatiin parannuksia erottelevien mallien oppimiseen rakenteellisesta anturitiedosta sekä erityisesti uusia menettelyjä tilannetiedon tunnistamiseen
Multiple Model Approaches to Modelling and Control,
No full textWhy Multiple Models?This book presents a variety of approaches which produce complex models or controllers by piecing together a number of simpler subsystems. Thisdivide-and-conquer strategy is a long-standing and general way of copingwith complexity in engineering systems, nature and human problem solving. More complex plants, advances in information technology, and tightened economical and environmental constraints in recent years have lead topractising engineers being faced with modelling and control problems of increasing complexity. When confronted with such problems, there is a strongintuitive appeal in building systems which operate robustly over a wide range of operating conditions by decomposing them into a number of simplerlinear modelling or control problems, even for nonlinear modelling or control problems. This appeal has been a factor in the development of increasinglypopular `local' and multiple-model approaches to coping with strongly nonlinear and time-varying systems.Such local approaches are directly based on the divide-and-conquer strategy, in the sense that the core of the representation of the model or controlleris a partitioning of the system's full range of operation into multiple smaller operating regimes each of which is associated a locally valid model orcontroller. This can often give a simplified and transparent nonlinear model or control representation. In addition, the local approach has computationaladvantages, it lends itself to adaptation and learning algorithms, and allows direct incorporation of high-level and qualitative plant knowledge into themodel. These advantages have proven to be very appealing for industrial applications, and the practical, intuitively appealing nature of the framework isdemonstrated in chapters describing applications of local methods to problems in the process industries, biomedical applications and autonomoussystems. The successful application of the ideas to demanding problems is already encouraging, but creative development of the basic framework isneeded to better allow the integration of human knowledge with automated learning. The underlying question is `How should we partition the system - what is `local'?'. This book presents alternative ways of bringing submodels together,which lead to varying levels of performance and insight. Some are further developed for autonomous learning of parameters from data, while others havefocused on the ease with which prior knowledge can be incorporated. It is interesting to note that researchers in Control Theory, Neural Networks,Statistics, Artificial Intelligence and Fuzzy Logic have more or less independently developed very similar modelling methods, calling them Local ModelNetworks, Operating Regime based Models, Multiple Model Estimation and Adaptive Control, Gain Scheduled Controllers Heterogeneous Control,Mixtures of Experts, Piecewise Models, Local Regression techniques, or Tagaki-Sugeno Fuzzy Models}, among other names. Each of these approacheshas different merits, varying in the ease of introduction of existing knowledge, as well as the ease of model interpretation. This book attempts to outlinemuch of the common ground between the various approaches, encouraging the transfer of ideas.Recent progress in algorithms and analysis is presented, with constructive algorithms for automated model development and control design, as well astechniques for stability analysis, model interpretation and model validation
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
- …
