Summaries for week 10
Physiological Measures of Presence in Stressful Virtual Environments
Since Virtual Environments are aimed at providing a sense of presence, their efficiency can be evaluated by measuring physiological responses of the body to the environment, such as heart rate, skin temperature or skin conductance. The method discussed here will be the measurement of heart rate since it is the most reliable and correlated one.
The virtual environment contains a training room and a pit room, the latter one having only a walkway and being 20 feet above another room. The aim is to carry balls while staying on the walkway. Thus, subjects showed symptoms revealing a situation of stress as if they faced the hole for real. Three other experiments involved the same experiment with a book, being on the edge of the wooden ledge and varying frame rates. Naturally, the sense of presence was improved by the physical sense of danger and a higher frame rate.
However, when subjects entered the environment several times, the heart rate was lower than the first time, which should not be the case in real life. But experiments showed that these measurements are reliable and valid. Furthermore, the measurements are very close for each testers, and validated all the hypothesis made so far.
But the system is not perfect and could be enhanced to provide a stronger sense of presence, by decreasing latency, allowing interactions with everything, adding physical cues and building self-avatars. So the aim is to build virtual environments providing exactly the same sensations than in real, but is it possible or will it never be reached?
Variations In Physiological Responses Of Participants During Different Stages Of An Immersive Virtual Environment Experiment
Most virtual environments are aimed at providing shocking experiences, therefore here the goal is to measure physical responses during a street scene to define the importance of this environment. One of the parameters is the realism of the people walking in the street.
The aim of a virtual environments is for the user to feel present within it. The heart rate and its variability is the most accurate measurement to determine it. To measure it, the experiment contains 8 characters in a street acting naturally, i.e. crossing street and avoiding collisions. These characters are whether cartoon-like, whether realistic. 40 subjects were used for this experience, which lasted 4 minutes. Before the experiment, a questionnaire and a training room were provided.
Since this environment is new to testers, they were stressed at the beginning, but their mental and physical stresses decreased quickly as they became familiar with it. However, the stresses increased when the subjects had to move. Also, users were more stressed while being immersed in a more realistic environment.
These results show that in virtual environments, users should have the time to become familiar with the tool before really using it. Furthermore, it would allow them to spot problems, increasing their confidence. But another possibility could be to play with the stress of the users, which could be a good immersion for certain applications.
The Uncanny Valley: Effect Of Realism On The Impression Of Artificial Human Faces
One of the aims in building a robot is to give a human face. Indeed, people enjoy it more. However, a very realistic robot with some defects is less enjoyable than a quite realistic robot, which is called the uncanny valley. The goal of this paper is to prove this theory.
The experiment developed was a web page showing a morphing face from a robot to a real face. Subjects had a grade to give to each face, whether it is pleasant or not. The first experiment displayed from 0% of realism to 100%. The second experience increased the range of realism, while the third one increased for each face the eyes up to one and half time their original size. Finally, the fourth experience used perfectly realistic faces instead of real faces.
Starting from 70% of realism, all subjects of experiment 1 had the sam pleasantness and did not prove the existence of the uncanny valley. The conclusion is the same for experiment 2. The third experiment showed that big eyes are not negative for artificial faces, but are very unpleasant for more realistic faces, leading to a graph showing a kind of uncanny valley. Finally, experiment 4 lead to the same observations.
Thus, these experiments prove the existence of the uncanny valley, but only when realistic faces have non-normal elements. But other experiments should involve videos and not only static images since motion is also an important factor. The final question to be asked is: Which features do a human eye analyze on a face?
Comments are closed.