
The Sense of Embodiment

Using a virtual reality headset, individuals can explore and move around in a virtual world designed entirely by computer. To interact with this environment, virtual hands are required to be able to grasp and handle objects. When users control a complete virtual body, called an avatar, a psychological phenomenon occurs: the sense of embodiment. This means that users are under the impression that the avatar is their own body. This effect can be created by synchronizing the user’s movements with those of the avatar: if the user moves a hand to grasp a virtual object, the avatar’s hand moves in the same way. The embodiment effect is even stronger when users see through the avatar’s eyes. It gives them the impression of being in its place, able to control it, and feeling the avatar’s actions as if they were their own.
Many studies have shown that modifying the avatar’s appearance can redefine a user’s self-image and modify certain psychological behaviors. Our study seeks to understand how our brain adapts to an avatar with extreme morphological deformation, and how this affects movement. More specifically: how do users plan and execute movements if one leg is twice as long as normal? Will they adjust the way they distribute their weight, or change the way they walk?
Gait Initiation Experience

To answer these questions, we conducted an experiment with twenty participants aged between 18 and 40. After donning a virtual reality headset, they could see a virtual copy of our laboratory and the reflection of their avatar in a virtual mirror. Next, three different conditions were applied to their virtual legs: either both legs were of normal size, or the leg taking the first step was lengthened, or the leg remaining on the ground was lengthened. In the latter two conditions, leg extension was progressive and located to the tibia segment. Participants were then asked to perform ten gait initiations, taking two steps forward. Gait initiation is a complex movement requiring whole-body muscle coordination and the integration of many sensory signals to move from standing to walking. As a result, gait initiation is likely to be disrupted by a virtual leg deformity.
We analyzed participant movements using cameras and markers placed on their bodies, and measured forces exerted on the ground with force platforms. The observed variables included the body’s center of gravity and center of pressure. The latter corresponds to the forces exerted by the feet on the ground, at different times of gait initiation.
Results showed that participants put less weight on their deformed leg, even before they started walking. Depending on whether the deformed leg was the one moving forward or the one remaining on the ground, this affected their stability differently when planning and executing movements. Interestingly, this uncertainty was similar to that of hemiparetic patients—who have weaknesses on one side of the body—in relation to their affected leg. One potential application of this study would be to create avatars with deformations adapted to specific deficits of hemiparetic patients, and help them compensate for these deficits. A rehabilitation program could include virtual reality sessions where patients—without being fully aware of it—would begin to progressively move their affected limb.
Additional Information
For more information on this research, please refer to the following paper: V. Vallageas, R. Aissaoui, I. Willaert, et D. R. Labbé, “Embodying a self-avatar with a larger leg: its impacts on motor control and dynamic stability,” IEEE Trans. Vis. Comput. Graph., vol. 30, no 5, p. 2066 2076, mai 2024, doi: 10.1109/TVCG.2024.3372084