Location but not amount of stimulus occlusion influences the stability of visuomotor coordination

Human unintentional and intentional interpersonal coordination in interaction with a humanoid robot

In order to establish natural social synchrony between two humans, two requirements need to be fulfilled. First, the coupling must be bi-directional. The two humans react to each other's actions. Second, natural social bodily synchronization has to be intentional or unintentional. Assuming that these essential aspects of human-human interactions are present, the present paper investigates whether similar bodily synchrony emerges between an interacting human and an artificial agent such as a robot. More precisely, we investigate whether the same human unintentional rhythmic entrainment and synchronization is present in Human Robot Interaction (HRI). We also evaluate which model (e.g., an adaptive vs non adaptive robot) better reproduces such unintentional entrainment. And finally, we compare interagent coordination stability of the HRI under 1) unidirectional (robot with fixed frequency) versus bidirectional (robot with adaptive frequency) rhythmic entrainment and 2) human intentional versus unintentional coupling. Fifteen young adults made vertical arm movements in front of the NAO robot under five different conditions of intentional/unintentional and unidirectional/bidirectional interactions. Consistent with prior research investigating human-human interpersonal coordination, when humans interact with our robot, (i) unintentional entrainment was present, (ii) bi-directional coupling produced more stable in-phase un-intentional and intentional coordination, (iii) and intentional coordination was more stable than unintentional coordination. To conclude, this study provides a foundation for modeling future social robots involving unintentional and bidirectional synchronization-aspects which seem to enhance humans' willingness to interact with robots.

Effects of unintentional coordination on attentional load

The goal of this study was to evaluate the impact of unintentional (spontaneous) coordination on high attentional visual load. More precisely, we wondered whether such coordination could free up some attentional resources and help improve performance in other more demanding attentional tasks. An experiment was performed in which participant attentional allocation was challenged by performing three tasks simultaneously while simultaneously being induced to unintentional entrain to an environmental rhythm. The first task was an interception task associated with a Stroop test to increase their attentional load. The second task was a reaction time test to alarms in different modalities (auditory, visual and bimodal) which was used to assess participant attentional load. The third task was a motor task in which participants were asked to swing their legs at a preferred frequency. The interface background brightness intensity was either synchronized in real time using a bidirectional coupling to participant leg movement or the background brightness was not changing at all. Our results on the reaction time task demonstrated that participants exhibited better reaction times for alarms in the bimodal condition than in the auditory condition and lastly for the visual condition. Also, participants exhibited a lower reaction time to alarms when the background brightness was synchronizing with their leg regardless the alarm modality. Overall, our study suggests a beneficial effect of unintentional environmental coordination on attentional resource allocation and highlights the importance of bidirectionality in interaction.

Timing and correction of stepping movements with a virtual reality avatar

Research into the ability to coordinate one's movements with external cues has focussed on the use of simple rhythmic, auditory and visual stimuli, or interpersonal coordination with another person. Coordinating movements with a virtual avatar has not been explored, in the context of responses to temporal cues. To determine whether cueing of movements using a virtual avatar is effective, people's ability to accurately coordinate with the stimuli needs to be investigated. Here we focus on temporal cues, as we know from timing studies that visual cues can be difficult to follow in the timing context. Real stepping movements were mapped onto an avatar using motion capture data. Healthy participants were then motion captured whilst stepping in time with the avatar's movements, as viewed through a virtual reality headset. The timing of one of the avatar step cycles was accelerated or decelerated by 15% to create a temporal perturbation, for which participants would need to correct to, in order to remain in time. Step onset times of participants relative to the corresponding step-onsets of the avatar were used to measure the timing errors (asynchronies) between them. Participants completed either a visual-only condition, or auditory-visual with footstep sounds included, at two stepping tempo conditions (Fast: 400ms interval, Slow: 800ms interval). Participants' asynchronies exhibited slow drift in the Visual-Only condition, but became stable in the Auditory-Visual condition. Moreover, we observed a clear corrective response to the phase perturbation in both the fast and slow tempo auditory-visual conditions. We conclude that an avatar's movements can be used to influence a person's own motion, but should include relevant auditory cues congruent with the movement to ensure a suitable level of entrainment is achieved. This approach has applications in physiotherapy, where virtual avatars present an opportunity to provide the guidance to assist patients in adhering to prescribed exercises.