Birgit Endrass

Natural interaction with culturally adaptive virtual characters

Recently, the verbal and non-verbal behavior of virtual characters has become more and more sophisticated due to advances in behavior planning and rendering. Nevertheless, the appearance and behavior of these characters is in most cases based on the cultural background of their designers. Especially in combination with new natural interaction interfaces, there is the risk that characters developed for a particular culture might not find acceptance when being presented to another culture. A few attempts have been made to create characters that reflect a particular cultural background. However, interaction with these characters still remains an awkward experience in particular when it comes to non-verbal interaction. In many cases, human users either have to choose actions from a menu their avatar has to execute or they have to struggle with obtrusive interaction devices. In contrast, our paper combines an approach to the generation of culture-specific behaviors with full body avatar control based on the Kinect sensor. A first study revealed that users are able to easily control an avatar through their body movements and immediately adapt its behavior to the cultural background of the agents they interact with.

Planning Small Talk behavior with cultural influences for multiagent systems

There are several factors that influence communicative behavior, such as gender, personality or culture. As virtual agents interact in a more and more human-like manner, their behavior should be dependent on social factors as well. Culture is a phenomenon that affects ones behavior without one realizing it. Behavior is thus sometimes perceived as inappropriate because there is no awareness of the cultural gap. Thus, we think cultural background should also influence the communication behavior of virtual agents. Behavioral differences are sometimes easy to recognize by humans but still hard to describe formally, to enable integration into a system that automatically generates culture-specific behavior. In our work, we focus on culture-related differences in the domain of casual Small Talk. Our model of culture-related differences in Small Talk behavior is based on findings described in the literature as well as on a video corpus that was recorded in Germany and Japan. In a validation study, we provide initial evidence that our simulation of culture-specific Small Talk with virtual agents is perceived differently by human observers. We thus implemented a system that automatically generates culture-specific Small Talk dialogs for virtual agents.

From observation to simulation: generating culture-specific behavior for interactive systems

In this article we present a parameterized model for generating multimodal behavior based on cultural heuristics. To this end, a multimodal corpus analysis of human interactions in two cultures serves as the empirical basis for the modeling endeavor. Integrating the results from this empirical study with a well-established theory of cultural dimensions, it becomes feasible to generate culture-specific multimodal behavior in embodied agents by giving evidence for the cultural background of the agent. Two sample applications are presented that make use of the model and are designed to be applied in the area of coaching intercultural communication.

Too close for comfort?: adapting to the user's cultural background

The cultural context of the user is a largely neglected aspect of human centered computing. This is because culture is a very fuzzy concept and even with a computational model of culture it remains difficult to derive the necessary information to recognize the users cultural background. Such information is only indirectly available and has to be derived from the observable multimodal behavior of the user. We propose the usage of a dimensional model of culture that allows applying computational methods to derive a users cultural background and to adjust the systems behavior accordingly. To this end, a Bayesian network is applied to allow for the necessary inferences despite the fact that the given knowledge about the users behavior is incomplete and unreliable.

Individualized agent interactions

Individualized virtual agents can enhance the users perception of a virtual scenario. However, most systems only provide customization for visual features of the characters. In this paper, we describe an approach to individualizing the non-verbal behavior of virtual agents. To this end, we present a software framework which is able to visualize individualized non-verbal behavior. For demonstration purposes, we designed four behavioral profiles that simulate prototypical behaviors for differences in personality and gender. These were then tested in an evaluation study.

Creating standardized video recordings of multimodal interactions across cultures

Trying to adapt the behavior of an interactive system to the cultural background of the user requires information on how relevant behaviors differ as a function of the users cultural background. To gain such insights in the interrelation of culture and behavior patterns, the information from the literature is often too anecdotal to serve as the basis for modeling a systems behavior, making it necessary to collect multimodal corpora in a standardized fashion in different cultures. In this chapter, the challenges of such an endeavor are introduced and solutions are presented by examples from a German-Japanese project that aims at modeling culture-specific behaviors for Embodied Conversational Agents.

Investigating culture-related aspects of behavior for virtual characters

In this paper, culture-related behaviors are investigated on several channels of communication for virtual characters. Prototypical behaviors were formalized in computational models based on a literature review as well as a corpus analysis, exemplifying the German and Japanese cultures. Therefore, aspects of verbal behavior, communication management and nonverbal behavior were taken into account. In evaluation studies conducted in the targeted cultures, each aspect’s impact on human observers was tested. With it, we investigated for which of the aspects, observers prefer agent behavior that was designed to resemble their own cultural background.

Time-Pie visualization: Providing Contextual Information for Energy Consumption Data

In recent years a growing number of information visualization systems have been developed to assist users with monitoring their energy consumption, with the hope of reducing energy use through more effective user-awareness. Most of these visualizations can be categorized into either some form of a time-series or pie chart, each with their own limitations. These visualization systems also often ignore incorporating contextual (e.g. weather, environmental) information which could assist users with better interpretation of their energy use information. In this paper we introduce the time-pie visualization technique, which combines the concepts of timeseries and pie charts, and allows the addition of contextual information to energy consumption data.

Rapid prototyping of social group dynamics in multiagent systems

In this article we present an engineering approach for the integration of social group dynamics in the behavior modeling of multiagent systems. To this end, a toolbox was created that brings together several theories from the social sciences, each focusing on different aspects of group dynamics. Due to its modular approach, the toolbox can either be used as a central control component of an application or it can be employed temporarily to rapidly test the feasibility of the incorporated theories for a given application domain. This is exemplified by applying the toolbox to different applications.

Cultural behaviors of virtual agents in an augmented reality environment

This paper presents a pilot evaluation study that investigates the physiological response of users when interacting with virtual agents that resemble cultural behaviors in an Augmented Reality environment. In particular, we analyze users from the Arab and German cultural backgrounds. The initial results of our analysis are promising and show that users tend to have a higher physiological arousal towards virtual agents that do not exhibit behaviors of their cultural background.