Objectives

The main research objectives of the BEAR project are to:

1) provide more realistic models and simulations of interactions between pedestrians based on human visual perception.

Simulation of pedestrians’ interactions, especially for immersion in virtual reality environments, requires representing the participant’s body (avatar) in the virtual world as well as the other virtual humans populating the virtual environment. While it has been shown that the avatar representation has a great influence on user’s behavior, e.g. how we interact with objects (Argelaguet et al., 2016), or our implicit interpersonal attitudes (Peck et al., 2013), it is seldom considered in previous works on pedestrian interactions. Therefore, our goal is to explore how both the representation of users, and of the virtual humans they are interacting with during locomotors tasks, affect the interaction and participants’ sense of avatar embodiment. Importantly, as it has recently been shown that participants are sensitive to self-generated visual self-motion cues for identifying with an avatar (Schettler et al. 2018), we aim at working on the animation (i.e., movements) of the virtual humans who interact with users, and evaluating how the combination of movement and shape influence the interaction.

2) understand how people gather the relevant visual information to control their motion.

We aim at developing new technical platforms and experiments to analyze gaze behavior. Our goal is to identify which elements of the environment are fixated (attended) to provide insight about the relevant visual information used to control motion. Indeed, gaze is directed towards areas that maximize the level of information about the environment to move safely (Cinelli et al., 2009), and can be related with some characteristics of the interaction between two walkers (i.e., crossing order) (Cinelli et al., 2007). We also recently demonstrated that gaze is attracted toward pedestrians within a crowd with high risk of collision (Meerhoff et al., 2018). We also want to analyze the influence of the physical properties of the other walker on gaze and locomotor behaviors, which will directly benefit from the improvements in realism resulting of objective 1 to conduct experiments in more ecological situations.

3) understand how altered perception in specific population modifies interactions between pedestrians.

Previous models and simulations consider healthy young adults and indicate that during interactions between pedestrians, perception and action are strongly coupled. Any change in the visual perception or the cognitive processing of visual perception would affect the action performed and then the models would not fit the behavior well. Previous studies have considered changes in visuo-motor strategies in locomotor tasks using an environment with static obstacles across several specific populations such as kids (Hackney and Cinelli, 2013a) , older adults (Hackney and Cinelli, 2013b) or previously concussed athletes (Baker and Cinelli, 2014). However, little is known about the interaction with other pedestrians, which is a more challenging task. We aim at conducting new experiments with these specific populations to provide more generic models that consider the individual characteristics in the visual perception.

The scientific challenges consist in capturing all the complexity of the human interactions. To tackle these challenges, we develop an experimental approach which includes a user interacting in real and virtual environments. Virtual reality allows to accurately manipulating the variables identified in real conditions.

References

  • Argelaguet, F., Hoyet, L., Trico, M., Lécuyer, A. (2016) The Role of Interaction in Embodiment: The Effects of the Virtual Hand Representation. In the Proceedings of IEEE Virtual Reality 2016.

  • Baker, C. S., & Cinelli, M. E. (2014). Visuomotor deficits during locomotion in previously concussed athletes 30 or more days following return to play. Physiological reports, 2 (12).

  • Cinelli, M. E., Patla, A. E., & Allard, F. (2009). Behaviour and gaze analyses during a goal-directed locomotor task. The Quarterly Journal of Experimental Psychology, 62(3), 483-499.

  • Cinelli, M. E., & Patla, A. E. (2007). Travel path conditions dictate the manner in which individuals avoid collisions. Gait & posture, 26(2), 186-193.

  • Hackney, A. L., & Cinelli, M. E. (2013a). Action strategies used by children to avoid two vertical obstacles in non-confined space. Experimental brain research, 229(1), 13-22.

  • Hackney, A. L., & Cinelli, M. E. (2013b). Young and older adults use body-scaled information during a non-confined aperture-crossing task. Experimental brain research, 225(3), 419-429.

  • Meerhoff, L., Bruneau, J., Vu, A., Olivier AH., Pettré, J. (2018) Guided by gaze: Prioritization strategy when navigating through a virtual crowd can be assessed through gaze activity. Acta Psychologica, 190, 248-257.

  • Peck, T. C., Seinfeld, S., Aglioti, S. M., Slater, M. (2013). Putting yourself in the skin of a black avatar reduces implicit racial bias. Consc. and Cognition, 22(3):779–787

  • Schettler A, Holstead I, Turri J, Barnett-Cowan M (2018) Visual self-motion feedback and the sense of self. European Conference on Visual Perception (ECVP). Trieste, Italy