1,721,011 research outputs found
Animal-Computer Interaction (ACI): An analysis, a perspective, and guidelines
No full textAnimal-Computer Interaction (ACI)’s founding elements are discussed in relation to its overarching discipline Human-Computer Interaction (HCI). Its basic dimensions are identified: agent, computing machinery, and interaction, and their levels of processing: perceptual, cognitive, and affective. Subsequently, three seminal studies are discussed, the ACI community should be become acquainted with. Next, three guidelines are defined that could help ACI to gain further maturity. We close with a brief conclusion
Continuous Affect State Annotation Using a Joystick-Based User Interface: Exploratory Data Analysis
No full textThe DLR (German Aerospace Center) aims to assess user’s affective state in motion simulators. To facilitate this
goal, a joystick-based user interface was used to gather reports on user’s emotions. This user interface allowed
continuous annotations, while video clips were watched. In parallel, several physiological parameters (e.g.,
electrodermal activity, heart and respiration rate) were acquired to record affective responses. An exploratory
data analysis of the users’ ratings (incl. several visualizations) that unveils several interesting data patterns is
presented
Transfer Learning for Rodent Behavior Recognition
No full textMany behavior recognition systems are trained and tested on single datasets limiting their application to comparable datasets. While retraining the system with a novel dataset is possible, it involves laborious annotation effort. We propose to minimize the annotation effort by reusing the knowledge obtained from previous datasets and adapting the recognition system to the novel data. To this end, we investigate the use of transfer learning in the context of rodent behavior recognition. Specifically, we look at two transfer learning methods with two different approaches and examine the implications of their respective assumptions on synthetic data. We further illustrate their performance in transferring a rat action classifier to a mouse action classifier. The performance results in the transfer task are promising. The classification accuracy improves substantially with only very few labeled examples from the novel dataset
Behavioural biometric identification based on human computer interaction
Get PDFAs we become increasingly dependent on information systems, personal identification and profiling systems have received an increasing interest, either for reasons of personali- sation or security. Biometric profiling is one means of identification which can be achieved by analysing something the user is or does (e.g., a fingerprint, signature, face, voice). This Ph.D. research focuses on behavioural biometrics, a subset of biometrics that is concerned with the patterns of conscious or unconscious behaviour of a person, involving their style, preference, skills, knowledge, motor-skills in any domain. In this work I explore the cre- ation of user profiles to be applied in dynamic user identification based on the biometric pat- terns observed during normal Human-Computer Interaction (HCI) by continuously logging and tracking the corresponding computer events. Unlike most of the biometrics systems that need special hardware devices (e.g. finger print reader), HCI-based identification sys- tems can be implemented using regular input devices (mouse or keyboard) and they do not require the user to perform specific tasks to train the system. Specifically, three components are studied in-depth: mouse dynamics, keystrokes dynamics and GUI based user behaviour. In this work I will describe my research on HCI-based behavioural biometrics, discuss the features and models I proposed for each component along with the result of experiments. In addition, I will describe the methodology and datasets I gathered using my LoggerMan application that has been developed specifically to passively gather behavioural biometric data for evaluation. Results show that normal Human-Computer Interaction reveals behavioural information with discriminative power sufficient to be used for user modelling for identification purposes
Annotation automatique d'images à base de Phrases Visuelles
No full textDate de fin de rédaction 15 Mai 2010.This thesis aims to propose a general model for automatic image annotation in the context of image retrieval. Seeking images requires abstract symbolic representations of theit semantic content (words, concepts ...) to satisfy the users information needs. While many studies have aimed to define a machine learning process of visual descriptors extracted from image regions, issues related to choices and grouping of descriptive and discriminative regions of different object classes are less studied. Visual variations of obects of a class cause serious problems for annotating images by object classes. These variations are caused by several factors: changes in scale, rotation and chages in brightness, in addition to variations of shapes and colors proper to any given object. Our work also aims to minimize the negative impact of this phenomenon. In this work, the passage from visual signal to its meaning is defined based on an intermediate representation called "Visual Phrases". These Phrases represent sets of regions of interest grouped according to a predetermined topological criterion. A learning process can detect relationships between Visual Phrases and object classes. Several evaluations of this approach have been conducted on the VOC2009 corpus. The results show the significant imact of the mode of grouping of regions of interest, and that a grouping based on spatial relationships among these regions gives the best results in terms of average precision.Ce travail de thèse a pour objectif de proposer un modèle général d'annotation automatique d'images pour la recherche d'information.La recherche d'information sur les documents images nécessite des représentations abstraites symboliques des images (termes, concepts) afin de satisfaire les besoins d'information des utilisateurs. Si de nombreux travaux ont pour objectif de définir un processus d'apprentissage automatique sur des descripteurs visuels extraits des régions d'images, les questions liées aux choix et aux regroupements des régions descriptives et représentatives des différentes classes d'objets sont peu étudiées. Les variations visuelles des objets d'une classe donnée posent de sérieux problèmes pour l'annotation par classes d'objets. Ces variations sont causées par plusieurs facteurs : changements d'échelle, rotation et changements de luminosité, en sus de la variabilité de forme et de couleur propre à chaque type d'objet. Notre travail vise aussi à minimiser l'impact négatif de ce phénomène. Dans ce travail, le passage du signal au sens se fonde sur une représentation intermédiaire appelée "Phrases Visuelles" qui représentent des ensembles de régions d'intérêt regroupées selon un critère topologique prédéfini. Un processus d'apprentissage permet de détecter les relations entre les Phrases Visuelles et les classes d'objets. Ce modèle d'annotation a fait l'objet de nombreuses évaluations sur le corpus VOC2009. Les résultats obtenus montrent l'impact significatif du mode de regroupement des régions d'intérêt, et qu'un regroupement prenant en compte les relations spatiales entre ces régions donne des meilleurs résultats en terme de précision moyenne
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe 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
Get PDF“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
Exploring historical cemeteries as a site for technological augmentation
No full textTangible and embodied technologies can enrich cultural heritage sites. Their design requires a solid understanding of the specific site, the needs and interests of user communities and stakeholders. Many types of heritage sites have been studied by HCI researchers, however our work focuses on a little-known one: historical cemeteries. Here we describe some early investigations of how the physical and socio-cultural contexts influence potential design solutions for two historic cemeteries, despite of a seemingly similar setting
Appropriate Similarity Measures for Author Cocitation Analysis
Get PDFWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
- …
