Biomedical Engineering Reference
In-Depth Information
of large environments is quite small—generally much less than the facilitative effect
of idiothetic on transient knowledge of local environments.
Finally, it is worth noting that questions about the relative influence of various
sensory systems (and combinations of sensory information) on spatial knowledge are
closely related to the question of how and whether spatial information from different
sources is integrated or combined into a coherent and unitary percept of space. Recent
research has provided empirical support for the combination of visual information
with auditory (e.g., [
1
]), vestibular (e.g., [
14
]), and haptic information [
26
]. Most of
this work is consistent with a Bayesian model of sensory integration in which the
weights assigned to various sources of information are determined by estimates of
the source's relative precision [
10
,
17
,
68
].
Having discussed the influence of various sensory systems to the acquisition of
knowledge of real world environments, we now turn our attention to the spatial
information available to users of VEs. As suggested earlier, a primary difference
between the sensory contributions of real and virtual environments is that it is typical
for the latter to involve the degradation or elimination of various sources of sensory
information that are commonly available in the real world.
1.4.2 Sensory Contributions in Virtual Environments
As VE systems have proliferated over the last several decades, a variety of interfaces
have emerged, with similarly diverse combinations of spatial information available
to users. For example, users of the 1950s-era Sensorama [
40
]—a widely recognized
precursor to modern VE systems—sat in front of a stationary display that presented
moving stereoscopic images, but lacked any accompanying head or neck motion.
The Sensorama's seat was capable of moving and thus provided a bit of sensorimo-
tor information, but the users' feet and legs remained relatively stationary, lacking
proprioceptive and kinesthetic input. Surround audio provided pre-programmed spa-
tial auditory information, and fans created the illusion of wind blowing from a certain
direction across the users' skin, which could be used to indicate the direction of one's
movement. Even synthetic smells could inform the user of being near one location
versus another. Of course the Sensorama was not a true VE system inasmuch as users
could not actively navigate or interact with their environment, but rather passively
took in a pre-programmed experience. Nonetheless, this early system captures the
types of sensory limitations and tradeoffs that are inherent in all VE designs. In
this final section of our chapter, we discuss three fundamentally different types of
VE systems: (a) desktop; (b) CAVE (Cave Automatic Virtual Environment); and (c)
HMD-based systems. Examples of these types of systems are depicted in Fig.
1.1
.
For each general type of system, we summarize the types of sensory information
that are available as well as which sources are absent, degraded, or limited, and what
is known about performance differences as a result of these limitations on sensory
information. Finally, general use cases under which each of these systems excels or
suffers are considered.
