Search

The aim of this special session at Measuring Behavior 2012 is to look at engagement and ways to measure engagement in situations where users are not glued to their chair and keyboard, that is, in sensor-equipped environments that are able to perceive nonverbal interaction behavior. And, moreover, we focus on activities that are not necessarily aimed at performing a particular task in the most efficient way. Rather the focus is on enjoyment during performance and satisfaction during and after finishing the interaction.

In this article the complexity of the construct engagement and three theories on this topic are discussed. Csikszentmihalyi's theory of flow is taken as starting point for the measurement of engagement. The measurement of each of its eight aspects is discussed, including its pros and cons. Regrettably, no overarching computational model is available. This article ends with a concise discussion.

Ambient spaces are characterized by sensors, which allow detailed observation of visitors and inhabitants, and by displays and actuators that allow the system to react to the observed behavior. Starting from a general ambient space, we are working towards an interactive environment that invites and seduces people to interaction with each other and with the environment. We aim at “play‿ rather than at “game‿; the visitor has no obligation to interact with the system, nor is there any specific goal to reach or task to complete.

A method is proposed to validate Microsoft’s Kinect as a device and, hence, to enable low fidelity, unobtrusive, robust sensing of behavior. The Xsens MVN suit is proposed as the measurements’ ground truth. An overarching framework is introduced that facilitates a mapping of both devices upon each other. This framework includes a complete processing pipeline for both the Xsens and the Kinect data, recorded in parallel, which, at the end of both pipelines, are mapped upon reach other. Next, two strategies are presented that aim to interpret the data gathered. We close with a brief discussion on the pros and cons of the protocol proposed.

Binnen een gecontroleerd werklastexperiment schatten we in een gesimuleerde online situatie werklastniveau op basis van EEG spectral power, ERPs en een combinatie hiervan. Dit lukt voor 33 van de 35 proefpersonen met gemiddelden van 80-90% correct.

This paper aims to give a brief overview of the current state-of-the-art in automatic measurement of affect signals in dimensional and continuous spaces (a continuous scale from -1 to +1) by seeking answers to the following questions: i) why has the field shifted towards dimensional and continuous interpretations of affective displays recorded in real-world settings? ii) what are the affect dimensions used, and the affect signals measured? and iii) how has the current automatic measurement technology been developed, and how can we advance the field?

In this paper we present the results of a pilot study investigating the impact of stress and cognitive load on the perceived interaction quality of a multimodal dialogue system for crisis management. Four test subjects interacted with the system in four differently configured trials aiming to induce low/high levels of stress and cognitive load. To measure the level of stress and cognitive load physiological sensors and subjective ratings were collected. After each trial the subjects filled in an evaluation questionnaire regarding the system interaction quality. In the end we conducted an in-depth interview with each subject. The trials were recorded with a webcam to facilitate the behaviour analysis. Results showed that both factors have an influence on the way subjects perceived the interaction quality, whereas the cognitive load seems to have a higher impact. Further quantitative experiments are needed in order to validate the results and quantify the weight of each factor.

This paper presents an overview of the work carried out at the HMI group of the University of Twente in the domain of multi-party interaction. The process from automatic observations of behavioral aspects through interpretations resulting in recognized behavior is discussed for various modalities and levels. We show how a virtual meeting room can be used for visualization and evaluation of behavioral models as well as a research tool for studying the effect of modified stimuli on the perception of behavior.

The creation of large, richly annotated, multimodal corpora of human interactions is an expensive and time consuming task. Support from annotation tools that make the annotation process more efficient is required, especially if the annotation effort involves really large amounts of data. Therefore we investigated how different properties of specific annotation tasks can have an impact on the design of a tool focused on that general class of tasks. In this paper we present our view on the considerations that should drive the design of new tools geared to specific tasks. The main dimensions that we consider are: observation vs interpretation, explicit and implicit input layers, segmentation, feedback, constraints, relations and the content of the annotation elements.

We present work on Obie, an embodied conversational agent framework. An embodied conversational agent, or talking head, consists of three main components. The graphical part consists of a face model and a facial muscle model. Besides the graphical part, we have implemented an emotion model and a mapping from emotions to facial expressions. The animation part of the framework focuses on the combination of different facial movements temporally. In this paper we propose a scheme of combining facial movements on a 3D talking head.

'Welzijn bestaat niet', is een gevleugelde uitdrukking van één van mijn collega's. Deze weliswaar ietwat ongenuanceere stelling geeft het probleem weer waarmee een gedragsbioloog in het welzijnsonderzoek geconfronteerd wordt: welzijn is niet direct meetbaar. Want de aanname is dat een dier zelf uitmaakt hoe het met zijn welzijn gesteld is. Dat hangt af van de omstandigheden waarin het dier zich bevindt. Maar hoe komt een etholoog te weten hoe een dier de eigen situatie beoordeelt? Je kunt er niet naar vragen. Nou ja, je kunt er wel naar vragen, maar een antwoord zul je niet krijgen. Of misschien toch wel? Het promotie-onderzoek van Patrick Zimmerman richtte zich op de vraag of het welzijn van leghennen te meten is op basis van geluiden die ze maken

Animal welfare plays a major role in the societal and scientific debate about the housing of laying hens. In this thesis it is assumed that welfare is good when physiologically an animal functions well. It is also proposed that affective states or feelings motivate an animal to perform behaviours that ensure proper biological functioning. Thwarting or blocking these behaviours impairs functioning and welfare. Emotional expressions, such as vocalisations, could be useful in assessing the welfare state of an animal.There are indications that a state of frustration, induced by thwarting of behaviour, in laying hens is expressed through a specific vocalisation, thegakel-call.Before the gakel-call can be used as a tool to assess welfare two requirements have to be met: a) frustration is consistently expressed through the gakel-call and b) the intensity of frustration is reflected in the number of gakel-calls or qualitative characteristics.The experiments in this thesis show that frustration, resulting from the thwarting of feeding, drinking, dustbathing and nesting behaviour, is expressed through the gakel-call. The motivation to feed affects the frustration response and is reflected in a higher number of gakel-calls. Social factors such as an audience-effect and social facilitation influence the gakel-call. The functional significance of these latter findings is discussed.As yet the gakel-call seems a useful parameter to study the impact of certain elements of a housing system on the frustration response of laying hens. Frustration can be an indicator of the welfare state of an animal and with the use of the gakel-call welfare of laying hens in different housing systems can be assessed.