Search

HCII 2021: 12th International Conference on Human-Computer Interaction (July 24-29, 2021), Item does not contain fulltext, Natural human interaction involves the fast-paced exchange of speaker turns. Crucially, if a next speaker waited with planning their turn until the current speaker was finished, language production models would predict much longer turn transition times than what we observe. Next speakers must therefore prepare their turn in parallel to listening. Visual signals likely play a role in this process, for example by helping the next speaker to process the ongoing utterance and thus prepare an appropriately-timed response. To understand how visual signals contribute to the timing of turn-taking, and to move beyond the mostly qualitative studies of gesture in conversation, we examined unconstrained, computer-mediated conversations between 20 pairs of participants while systematically manipulating speaker visibility. Using motion tracking and manual gesture annotation, we assessed 1) how visibility affected the timing of turn transitions, and 2) whether use of co-speech gestures and 3) the communicative kinematic features of these gestures were associated with changes in turn transition timing. We found that 1) decreased visibility was associated with less tightly timed turn transitions, and 2) the presence of gestures was associated with more tightly timed turn transitions across visibility conditions. Finally, 3) structural and salient kinematics contributed to gesture's facilitatory effect on turn transition times. Our findings suggest that speaker visibility - and especially the presence and kinematic form of gestures - during conversation contributes to the temporal coordination of conversational turns in computer-mediated settings. Furthermore, our study demonstrates that it is possible to use naturalistic conversation and still obtain controlled results.

One of the biggest challenges in Group Therapy is to track each patient’s experience and feeling without him/her noticing. Altering the familiarity of the mutual support group routine may weaken the therapeutic efficacy of the intervention. It must be avoided the “Elephant in the room’s Effect”: everyone knows is being observed and acts consequently. Therapists struggle and spend years of training on developing the skills they need to “silently” monitor all patients at the same time. From our perspective, we wonder whether and how technology can be a support for therapists in such a challenging task. More precisely, how to provide them with a non-invasive support tool that is invisible to the end-users, but at the same time ever-present for the caregivers. Basically, we asked ourselves: Can we deceive “the Elephant in the room”? Therapists may benefit from automatic measures indicating how the participants perceive the session and gathering the participants’ feedback is one path to develop valuable mutual support interventions. Our work describes the design, development and assessment of a non-invasive tool to monitor a Group Session.

"null", Item does not contain fulltext, Chatbots are typically used in dialogue systems for various purposes such as customer service and information acquisition. This paper explores enhancement of social skills training for security guards with the use of chatbots. More specifically, we designed a chatbot using text and voice as input to study the acceptance and the impact of the system to training security guards in deal with stress situations. The result of a pilot experiment and a survey are presented and discussed. Finally, we discuss possible improvements and future work.

19th International Conference on Human-Computer Interaction, HCI International 2017, Vancouver, BC, Canada, July 9-14, 2017, Item does not contain fulltext, This paper presents an algorithm to automatically detect interpersonal stance in vocal signals. The focus is on two stances (referred to as 'Dominant' and 'Empathic') that play a crucial role in aggression de-escalation. To develop the algorithm, first a database was created with more than 1000 samples from 8 speakers from different countries. In addition to creating the algorithm, a detailed analysis of the samples was performed, in an attempt to relate interpersonal stance to emotional state. Finally, by means of an experiment via Mechanical Turk, the performance of the algorithm was compared with the performance of human beings. The resulting algorithm provides a useful basis to develop computer-based support for interpersonal skills training.

19th International Conference on Human-Computer Interaction, HCI International 2017, Vancouver, BC, Canada, July 9-14, 2017, Item does not contain fulltext, A Virtual Reality based training can be an interesting method to teach crowd managers and emergency responders how to act in emergency situations under pressure. Compared to watching Two-Dimensional Video, Virtual Reality is assumed to induce stronger emotions and a more real-life experience of the emergency situation. To test this assumption, sixty participants were tested on whether there is a difference in inducing fear between cardboard Virtual Reality glasses and watching a 2D video. Subjective (PANAS) and objective (heart rate, blood pressure) ratings of their mood were measured before, during and after watching a horror movie. Results show that both VR and 2D induced fear significantly in all subjective and objective measures, but that there was no significant difference between these video screen conditions. Based on inducing fear alone, a video based training could be as effective as a cardboard VR training. More research is needed to investigate this further.

This paper reports on the current field of narrative-based game design through case study analysis with a particular focus on balancing high narrative agency with low production resources.

Global producers of mobile communication devices recognize the importance of cultural differences in the emerging markets; however it seems that the main concentration in both academic and business areas is on the large number of users with low incomes, while users from other classes of these societies are not studied well. In this study after set of integrated reviews on areas of Mass Customization, New Product Development and Mobile HCI an experiment was planned based on the unexplored aspects of users' culture and mobile communication devices relationships. A number of young educated users from middle class tested a new smart phone during its marketing process in Iran. They were sampled based on a global producer's marketing program. After a phase of self documentation, users selected two applications of the device for the usability tests and found culture related usability problems during the tests. Finally they proposed solutions in a participatory design process.