|
Technion: Subliminal cues in haptics A famous phenomenon in visual perception is known as subliminal perception. Subjects are not aware of seeing an object, yet behavior indicates that subjects have some knowledge of the object, although they testify of no perception. Technion investigated if there is a similar phenomenon in virtual haptic feedback, and therefore could be exploited for use in IMMERSENCE. Using an immersive virtual environment that comprises haptic and visual interaction, Technion run a texture-difference- recognition test in which subjects glide a pen-like stylus along a varying rough surface. The experiment was carried out for several values of the friction coefficient, and several values of difference in friction coefficient between the two surfaces. The researchers found that below a threshold limit of roughness, although subjects claimed that texture was uniform when it was not. Still, subjects changed the applied normal force, adapting to the new (unrecognized) value of roughness. That is, they did not recognize changes in the haptic sensory cues, but behaved as if they did. This suggests that haptic perception may be of two kinds: aware/unaware. A threshold value indicates the borders of aware unaware perception. Below this threshold, subjects are unaware of haptic perception and of the affects on behavior. Above this threshold, subjects are aware of perception and of affects on behavior. In many manual tasks such as surgery, product assembly, palpating a lump, feeling the surface skin for imperfections, sculpting, piano playing, fabric design, etc, touch cues are essential for high quality performance. The task quality performance is closely related to tactile sensitivity. Unaware changes in input stimuli can affect people's behavior (subliminal perception). The methodology Technion developed improves the reliability of the methods commonly used, by adding an objective tool of analysis to object's reactions to stimulus change, thanks to the recording capabilities of our experimental device. Experimtenal set-up and results in detail: The VR device used at Technion allows to perform a set of measurements concerning the haptic interaction. The figure below depicts the actual view and schematic representation of the virtual environment that the participant sees during the experiment. The VR setup utilized, in fact, allowed to design VR objects with arbitrary physical properties. The researchers used the Reachin function that simulates something akin to sand paper. The haptic device's motion is constantly stopped, and a new starting friction value is randomly calculated using a Gaussian probability distribution with a mean and standard deviation supplied by the team. The two values that can be altered are the mean and the standard deviation. The input in the mean value is the friction coefficient. The input in the standard deviation allows the randomness to be contained, and it was kept constant throughout the whole experiment.
Visual stimuli consisted of a rectangular shaped surface, twenty centimeters long, with programmable roughness qualities resembling sand paper. The surface is divided in two parts with different friction coefficients. The line that divides the two parts is randomly located in the rectangle's ten central centimeters to allow data analysis (see next subsection), and visually undetectable. The two circles and the straight line are only indicative and don't form part of the surface. Their meaning, and that of the ruler, will become evident in the next paragraph. For each run, the input consists of a pair of values of the friction coefficient. They are set so that their difference covers a large spectrum of values. For each set of runs, the lower value is kept constant, and it is always in the same side of the rectangle. The subject is instructed to move gently the force feedback handle from one circle to the other one, along the straight line across the surface, from the smoother side to the rougher side. S/he must report, by pressing the upper left button if s/he did not feel the roughness change, or the right button (see figure above), if s/he feels any roughness difference. Additionally, if s/he reports to feel the roughness change, s/he must press the ruler (see same figure) near the zone where in his mind, the change occurs. The distance from the actual point of change is recorded. The normal component of the force applied by the subject on the surface is measured as a function of the distance from the point of change in roughness. Pointing the ruler introduces an objective measurement of sensitivity, avoiding false reports of awareness of change in surface roughness. The researchers set a length of one centimeter as the maximal distance between the pressing point on the ruler and the actual point of change for accepting the subject's report sensing the change. Then they recorded the applied normal force to demonstrate that subliminal perception has happen. First, they needed to validate that in the VR environment a minute difference in the input values of the friction coefficient causes detectable changes in the applied normal force, even if the subject wasn't aware of the change. That is, subconscious, or subliminal perception toke place. The figure below shows examples of measurements of the applied normal force as a function of the distance from the point of change in roughness, when subliminal and conscious perception happened. The bumpiness in the measured applied normal force is due to the randomness in the value of the friction coefficient with a maximal probability near the inputted mean, due to the Gaussian distribution.
Change in applied normal force for subliminal and aware perception For every run, Technion calculated the average applied normal force, before and after the roughness change, neglecting the edges (start and end of movement). As a standard, they used for analysis the data recorded between four centimeters, before and after roughness change. Statistically they calculated the significant level in force change. Worth notifying, through all the runs Technion obtained p < 0.01. In the above examples the average measured force was 2.12 N and 2.56 N for subliminal perception, and 2.10 N and 2.62 N for non-subliminal perception, before and after the roughness change, respectively. For every basic value of the friction coefficient (S in second figure 2 above) they obtained the limit value of the other friction coefficient (S+Smin in Figure 3) where conscious perception becomes subliminal. Then, they calculated the Just Noticeable Difference, JND for each subject for each value of S applied in this experiment, using the JND's definition : JND =Smin/S Each subject performed thirty runs in three different basic friction coefficient regimes . An additional ten runs were performed for calibration, that is, null difference in the pair of friction coefficients that implies null difference in the normal applied force. Therefore each subject performs one hundred runs in our experiment. The three different dimensionless friction coefficients, µ, used as the smoother value in each set of runs (S in second figure), were: 0.1, 0.3, and 0.5, where µ is defined as: µ Ffriction/N. That is, the friction force between the surfaces in contact divided by their mutual normal force. Each regime was divided into ten pairs of friction coefficient values. So, for the same values difference the task was repeated three times. Below a summary of the possible output for every run: a) Subject feels the change à distance from change point < 1 cm à conscious perception. b) Subject feels the change à distance from change point > 1 cm à subliminal perception. c) Subject doesn't feel the change à subliminal perception. For every run, a change in the normal applied force must be detected. In the case that for a given friction coefficient difference, the subject was aware of the surface change (case a) in part of the runs, and in others he didn't (cases b or c), it was considered as subliminal perception when case a, happens in two of the three runs. The values of the friction coefficient in the different regimes, and the maximal distance from the point of change were chosen as suitable, after many pilot runs in our laboratory.
The table on the right shows the average change in the applied normal force for all 11 subjects, when roughness surface changes at the subliminal limit, for all three friction-coefficients regimes. It is interesting to note the similar values for all regimes and the relatively narrow standard deviation values, despite the large ones the researchers obtained for JNDs. The reason might be the smoothed and gentle hand movement across the virtual surface, that subjects was instructed to do.
As the table on the right shows, it is evident that when the stimulus threshold is approached, the calculated JND increases sharply, in accordance to literature. The large standard deviation for the 0.1 friction coefficient seems to imply that threshold sensitivity for this stimulus is highly individual. Technion found that subjects were not aware of texture roughness change below a threshold limit, yet the normal force they applied, changed. They didn't recognize changes in the sensory cues, but behaved as if they did. This suggests that performance can be affected and possibly improved through subliminal cues by reducing the cognitive load, allowing the transfer of additional information otherwise impossible. |
||||||||||||||||||||||||
|
Related deliverables: |
||||||||||||||||||||||||
| Related papers: | ||||||||||||||||||||||||
| >>back to overview |
