Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content

    Elisabeth André

    From 14.03.04 to 19.03.04, the Dagstuhl Seminar 04121 ``Evaluating Embodied Conversational Agents'' was held in the International Conference and Research Center (IBFI), Schloss Dagstuhl. During the seminar, several participants... more
    From 14.03.04 to 19.03.04, the Dagstuhl Seminar 04121 ``Evaluating Embodied Conversational Agents'' was held in the International Conference and Research Center (IBFI), Schloss Dagstuhl. During the seminar, several participants presented their current research, and ongoing work and open problems were discussed. Abstracts of the presentations given during the seminar as well as abstracts of seminar results and ideas are put together in this paper. The first section describes the seminar topics and goals in general. Links to extended abstracts or full papers are provided, if available.
    Understanding emotions of others requires a theory of mind approach providing knowledge of internal appraisal and regulation processes of emotions. Multi-modal social signal classification is insufficient for understanding emotional... more
    Understanding emotions of others requires a theory of mind approach providing knowledge of internal appraisal and regulation processes of emotions. Multi-modal social signal classification is insufficient for understanding emotional expressions. Mainly, because many communicative, emotional expressions are not directly related to internal emotional states. Moreover, the recognition of the expression's direction is neglected so far. Even if social signals reveal emotional aspects, the recognition with signal classifiers cannot explain internal appraisal or regulation processes. Using that information is one approach for building cognitive empathic agents with the ability to address observations and motives in an empathic dialogue. In this paper, we introduce a computational model of user emotions for empathic agents. It combines a simulation of appraisal and regulation processes with a social signal interpretation taking directions of expressions into account. Our evaluation show...
    Embodied conversational agents may take on a diversity of roles in learning and advisory scenarios including virtual teachers, advisors, learning companions and autonomous actors in educational role play. They promote learner motivation,... more
    Embodied conversational agents may take on a diversity of roles in learning and advisory scenarios including virtual teachers, advisors, learning companions and autonomous actors in educational role play. They promote learner motivation, engagement, and self-confidence, and may help prevent and overcome negative affective states of learners, such as frustration, and fear of failure. The chapter will provide guidelines and approved methods for the development of animated pedagogical agents including the extraction of multimodal tutorial strategies from human-human teaching dialogues as well as the simulation and evaluation of such strategies in computer-mediated learning environments.
    Summary The objective to develop more human-centered, personalized, and at the same time more entertaining interfaces immediately leads to the metaphor of an embodied conversational agent that employs gestures, mimics, and speech to... more
    Summary The objective to develop more human-centered, personalized, and at the same time more entertaining interfaces immediately leads to the metaphor of an embodied conversational agent that employs gestures, mimics, and speech to communicate with the human user. Looking at past and current projects, the current paper discusses an ongoing and manifold evolution of embodied conversational agents from conversational settings with single presenters to interactive performances where the user may participate both as an observer and a presenter. We report on new trends, such as the integration of characters in mixed realities as well as endeavours to endow characters with social behaviors.
    Research Interests:
    In this paper, we present an interactive eye gaze model for embodied conversational agents in order to improve the experience of users participating in Interactive Storytelling. The underlying narrative in which the approach was tested is... more
    In this paper, we present an interactive eye gaze model for embodied conversational agents in order to improve the experience of users participating in Interactive Storytelling. The underlying narrative in which the approach was tested is based on a classical XIXth century psychological novel: Madame Bovary, by Flaubert. At various stages of the narrative, the user can address the main character or respond to her using free-style spoken natural language input, impersonating her lover. An eye tracker was connected to ...
    Laughter and fillers like “uhm” and “ah” are social cues expressed in human speech. Detection and interpretation of such non-linguistic events can reveal important information about the speakers’ intensions and emotional state. The... more
    Laughter and fillers like “uhm” and “ah” are social cues expressed in human speech. Detection and interpretation of such non-linguistic events can reveal important information about the speakers’ intensions and emotional state. The INTERSPEECH 2013 Social Signals Sub-Challenge sets the task to localize and classify laughter and fillers in the “SSPNet Vocalization Corpus” (SVC) based on acoustics. In the paper at hand we investigate phonetic patterns extracted from raw speech transcriptions obtained with the CMU Sphinx toolkit for speech recognition. Even though Sphinx was used out of the box and no dedicated training on the target classes was applied, we were able to successfully predict laughter and filler frames in the development set with  87% accuracy (unweighted average Area Under the Curve (AUC)). By accumulating our features with a set of standard features provided by the challenge organizers results increased above 92%. When applying the combined set to the test corpus we achieved 87:7% as highest score, which is 4:4% above the challenge baseline.
    Research Interests:
    This chapter is from the forthcoming The Oxford Handbook of Affective Computing edited by Rafael Calvo, Sidney K. D'Mello, Jonathan Gratch, and Arvid Kappas. Although many papers emphasize the need to incorporate cultural values and norms... more
    This chapter is from the forthcoming The Oxford Handbook of Affective Computing edited by Rafael Calvo, Sidney K. D'Mello, Jonathan Gratch, and Arvid Kappas. Although many papers emphasize the need to incorporate cultural values and norms into emotional agent architectures, work that actually follows such an integrative approach is rare. To construct anthropomorphic agents that show culture-specific emotional behaviors, researchers must investigate how emotions are conveyed across cultures and how this knowledge can be used to tune emotion recognizers to a particular culture. Models of appraisal and coping have to be enriched by models of culture to simulate how the agent appraises events and actions and manages its emotions depending on its alleged culture. Finally, mechanisms are required to modulate the expressiveness of emotions by cultural traits to convey emotions with right level of intensity and force. Starting from work done in the agent research community, this paper discusses how existing work on equipping anthropomorphic agents with emotional behaviors can be extended by considering culture-specific variations.
    Research Interests:
    Research Interests:
    Research Interests:
    Research Interests:
    Google, Inc. (search). ...
    ... into one presentation unit because there might be crossreferences from text to graphics. ... of the Persona's behaviours, we have defined a declarative authoring language which represents ... a subjective... more
    ... into one presentation unit because there might be crossreferences from text to graphics. ... of the Persona's behaviours, we have defined a declarative authoring language which represents ... a subjective measure), and (2) its effect on the subject's comprehension of presentations ...

    And 30 more