Hardware/Software Issues

 

 

 

 

 

 

What are head mounted displays?

           Head mounted displays (HMDs) are stereoscopic devices that transfer two-dimensional visual images into three-dimensional visual images, and resemble oversized helmets (Schneiderman, 1998). HMDs generally cover 60 degrees vertically and 100 degrees horizontally and block out all other images in the immediate environment, so the user is viewing only what is directly in front of them on the screen. Coupled with synthetic sound, HMDs can substantially contribute to the user’s illusion of telepresence.  However, improperly designed HMDs can contribute to operator discomfort and visual stress. Most researchers agree that displays must approach real-time (under 100 milliseconds) in presenting images to users. Low-resolution displays are acceptable while users or the objects are moving, but when stationary, higher resolution is essential to create realism in the simulation. Technologically advanced HMDs have been developed recently by the University of Illinois, the University of North Carolina, and Boeing Computer Services (Schneiderman, 1998, Ellis, 1995). (Image courtesy of Visnet)

See also Field of View
See also Head Trackers

 

 

 

 

 

 

 

What is the current state of the art in head trackers?

            Head mounted displays (HMDs) can provide different viewpoints to users depending upon the position of the head. Head tracking is a technique that measures the position of the head and produces new images for users when the head is turned a certain way (Schneiderman, 1998). For example, upon looking left, the user could see a forest, but if he or she turns slightly to the right, the forest may give way to a river. Advanced head tracking and head sensing technologies require high sensor precision (within 1 degree) and rapid response (within 100 milliseconds). The current state of the art in head trackers is boom-mounted display such as the Fakespace BOOM3C (Binocular Omni-Orientation Monitor) that provides head sensing in addition to high-resolution visual displays. Another example is Ascension's Flock of Birds that provides fast stable head tracking in a number of military and commercial simulation products. Many such head trackers are mounted on the user’s head, but future technologies point to other embedding trackers in hats or eyeglasses. (Schneiderman, 1998)

  See also Field of Regard
See also Head Mounted Displays

   

 

 

 

How good are speech recognition devices?

 

            Designers of hardware have made continuous progress in speech and voice-recognition devices, with the ultimate goal of speaking to computers and listening to them speak to humans. Discrete-word recognition devices can recognize up to a 200-word vocabulary, but require training with an individual user in order to remain reliable (Schneiderman, 1998). Many advanced speech recognition devices have been used in military aircraft, medical operating rooms, and training laboratories. Unfortunately, although speech recognition has advanced, major problems still exist when used in these environments. Problems occur when background sounds change, when the user is under stress, and when words in the vocabulary sound alike (Schneiderman, 1998). Current research projects are improving recognition rates in such conditions, and eliminating the need for intensive individual user training. IBM’s new ViaVoice system demonstrates specialized vocabularies in such fields as radiology and promises simpler training, higher recognition rates, and a newly designed interface. In the future, designers are focusing on continuous speech recognition devices, which have vocabularies of up to 10,000 words and better recognize the boundaries between spoken words (Schneiderman, 1998).

 See also Speech Production

 

 

 

 

 

 

How good is speech production?

           Speech generation devices attempt to produce synthetic speech, with early research in speech generation dating back to the 18th century. Michaelis and Wiggins (1982, cited in Schneiderman, 1998) note that speech generation is optimal if the message is simple, short, requires an immediate response, and deals with events in time. Speech generation is also helpful for users who visual systems are already overloaded, or for users at high altitudes suffering from eyestrain problems. In other words, an automatically generated message or alert can be crucial to users who already have a high cognitive workload (Schneiderman, 1998). In low workload situations, however, users may be apt to miss the alerts, pay less attention to computer-generated speech, or find it generally annoying. Success stories with speech generation devices have included telephone based voice information systems for banking such as Fidelity Automated Service Telephone (FAST) and airline schedule information systems such as American Airlines Dial-AA-Flight, which have voice input/output features coupled with keypad options (Scheiderman, 1998).

See also Speech Recognition

 

 

 

 

What is lag? How can lag be measured? What is the recommended maximum lag allowed to prevent side effects?

           Lag problems are major concerns in simulation. Transport delay, also referred to as display lag, refers to the time lag between the moment at which the head moves and the moment in which the image is displayed (Schneiderman, 1998). Essentially, display lag represents the response of the device to head movement. Display lag can lead to image smearing or the illusion that the image is “swimming” on the screen during the start and finish of a head motion. It can also detract from the realism of the simulation, because users immediately sense that they are interacting with computer-generated images and become aware that they are not in a “real world” experience. More importantly, display lag can lead to the development of eyestrain problems, nausea, headaches and other symptoms of simulator sickness (Kolasinski, Goldberg, & Hiller, 1995).

            The measure of display lag is virtually the time taken to determine head orientation and display the corresponding image (Schneiderman, 1998). Display lag is measured in the order of milliseconds, and an undesirable display lag has been found to be greater than or equal to 20 milliseconds (Draper, 1996). It has been found that display lag leads to images that are positioned improperly, and this error increases with head movement velocity (Bryson & Fisher, 1990). Such display lags have been found to constraint a user from making quick head movements because he or she is disturbed by the improper positioning of images (Grunwald, Kohn, & Merhav, 1991). It also leads to decrements in performance and simulator sickness symptoms.

 

 See also Simulator Sickness

 See also Head Mounted Displays

   

 

 

 

 

 

 

 

 

 

What is stereoscopy? Do I need a binocular visual display? What is the difference between biocular and binocular displays?

            Humans perceive depth and distance using a phenomenon known as binocular parallax. This involves the use of the left and right eyes, each containing broad cones of perception and each seeing a slightly different image. The convergence of these two images results in one perceived image in three dimensions and provides a basis for judging distance (Wickens, 1992). The technology of stereoscopy involves using optical apparatus and pairs of images to create three-dimensional illusions (Rheingold, 1991).  Artificial ways of presenting stereoscopic images involves special glasses that show image pairs to the proper eyes, referred to as a binocular display. However, there are ways of presenting stereoscopic images without a binocular display with the use of lenticular lenses or biocular displays, which take vertical slices of the two images and interweave them together. Thus, both eyes view a single image source at one time. In reality, the left eye image is viewed with the left eye and right eye image is viewed with the right eye, resulting in a carefully crafted illusion of depth (Rheingold, 1991). 

 See also Stereoscopy/3D
See also Visual Perception

 

 

 

 

 

 

How do I measure visual resolution?

            Visual resolution is measured by the number of pixels contained on a display monitor or screen, and is expressed in terms of the number of pixels on the horizontal axis and the number on the vertical axis. The word pixel is derived from the combination of the words “picture” and “element”. By definition, it is the basic unit of programmable color on a computer display or in a computer image (Whatis.com, 2002). Typical resolution is 768 by 1024 pixels, but 1280 by 1024 pixels is also a common resolution size. Furthermore, the sharpness of the image on a display depends on the resolution and the size of the monitor and the characteristics of the human eye. The same pixel resolution will be sharper on a smaller monitor and gradually lose sharpness on larger monitors because the exact same number of pixels is being spread out over a larger distance (Whatis.com, 2002). (Image courtesy of National Aerospace Laboratory NLR)

See also Visual Perception
See also Binocular Visual Displays

 

 

 

 

 

 

Where do I find databases Models? Spatial databases?

          Simulation often involves using a computer-based model to convince the users that they are actually in a virtual environment (Rheingold, 1991). For example, if a simulation seeks to mimic airplane flight, a mathematical model of the wing is needed to see how it behaves mechanically when subjected to various tests. Modeling in this sense is an important cognitive tool for understanding how a device operates (Rheingold, 1991). General collections of mathematical models for developing simulations can be found in the U.S Army Modeling and Simulation Resource Repository and the U.S. Navy Modeling and Simulation Office. Similarly, databases provide a storehouse of information used to develop high fidelity military simulations. The Modeling and Simulation IAC provides databases in special interest areas including the online WARMOND database for special warfighter needs.  In addition, the National Library of Medicine provides additional databases for medically related models and simulations.

            A spatial database is a collection of spatially referenced data that acts as a model of reality for the simulation designer. Generally, geographic data in simulations is most commonly represented by spatial databases. Such databases involve the use of spatial operations like spatial joins, map overlays, nearest neighbor queries and others to accurately portray realistic objects in a simulation (Koperski, 2002). The National Standard for Digital Cartographic Databases specifies the standards for developing and using spatial databases (Nyerges, 1997). In addition, a list of contacts and projects for spatial database research is also available for designers and developers.

See also Modeling Human Figures

 

 

 

 

 

 How do I model human figures?

             Avatars, or representations of human beings, are crucial aspects of a realistic simulation. With that said, virtual human beings must look, speak, and move as logically and realistically as possible. Advanced research in human figure modeling is being done at the University of Pennsylvania.  At the University’s Center for Human Modeling and Simulation, two advanced modeling techniques have been developed. First, a software package called “Jack” provides a three dimensional environment for manipulating and controlling human figures. In addition, the center is also developing an advanced set of algorithms for simulating human arms and legs. This package is referred to as Inverse Kinematics using Analytical Methods (IKAN) and is available on the above mentioned website. Additional human modeling resources can be found within the Human and Organizational Behavior Modeling special interest area of the Modeling and Simulation Information Analysis Center. (Image courtesy of the University of Pennsylvania Center for Human Modeling and Simulation).

See also Avatar
See also Missing Team Members