CSE Research
Enhanced Immersive Virtual Environment Reorientation: Intelligent Imperceptible Alignment
Immersive virtual reality (IVR) is a self-recognition of being present in a virtual world that is generated by VR system. Immersive VR systems that utilize a head mounted display may be capable of supporting natural navigation of virtual worlds. In such systems, physical movements are mapped to movement in a virtual world. For instance, walking 10 meters forward in an area in which movement can be be tracked may result in an equivalent movement in the virtual world. In such a system, the virtual world may be larger than the tracking area. This can lead to users not being able to walk freely while exploring virtual worlds due to the possibility of leaving the tracking area and potentially colliding with obstacles such as walls. Previous researchers have developed resetting or reorientation techniques to solve this problem. The basic approach of resetting techniques is to instruct the user to stop translational movement and turn 360 degrees in virtual world whenever a obstacle in the physical world is approached. The virtual turn that is accompanied by the physical turn is either more or less than 360 degrees and is also in place.
Real-Time Prediction of Virtual Environment Motion Sickness Using Postural Sway
Researching causes of motion sickness, specifically motion sickness caused by simulators and virtual environments (VE), has been on the rise due to their increased use for training and entertainment purposes. There have been multiple theories presented and studied on the triggers and preventative measures of motion sickness. One approach examines a participant's postural stability, or how the person's posture sways, while subjected to a VE. Previous research has shown a relationship between the complexity of visual stimulants and the actions of the participant. This relationship differs for well and sick individuals. Traditionally, researchers perform pre and post experiment balance checks and questionnaires to determine if a person was feeling sick or not. However, very little research has been successful in detecting motion sickness, in real time, before it occurs. We propose an algorithm that continually analyzes a participant's postural stability and predicts their current sickness potential while they are subjected to a simulation or VE. Using this algorithm, we hope to provide early notification of sickness before major symptoms of sickness occur.
Miami Illinois Language User Extensible Dictionary
In this project we develop a digital dictionary for the Miami Illinois language that is configurable and extensible by the end-user. The four (4) primary goals of the project are: (1) To implement a new dictionary designed to reflect grammatical features of the Myaamia language; (2) To improve the search capability required by the user community; (3) To implement web browser and iOS versions of the dictionary; and (4) To identify and implement a design framework that allows language researchers to extend the dictionary to support language building blocks as the language is reconstructed and revitalized.
Artificial Potential Field Redirected Walking
Many virtual environment (VE) systems allow users to navigate within virtual worlds through natural walking in a pre-defined area within which their physical position is tracked. Since users cannot move beyond the borders of their physical tracking area, previous work has established approach to virtually extend that space called redirected walking (RDW). RDW strategies slowly rotate the VE about the head of the users which subconsciously alter their heading to follow. In this way we can direct users away from walls while they retain choice within the VE. RDW strategies differ in the direction and amount the virtual world is rotated each frame. We propose a generalized RDW strategy called artificial potential field (APF) that turns the user toward a target movement direction from on a simulated-force vector field mapped onto the physical tracking area. The vector field approach gives us the flexibility to take the shape of the tracking area into account and easily supports multiple users in the same tracking area simultaneously. We present simulation data to evaluate the performance of our APF strategy. Compared to previous approaches, single-user APF increased the number of meters walked per wall contact by 28%. In multi-user scenarios, APF reduced the number of potential collisions between immersed users by 31 ± 4% (95% confidence interval) compared to control.
Undergraduate Research
The Department of Computer Science & Software Engineering offers many opportunities for undergraduate research with faculty. CSE 491 is an undergraduate research course that allows you to work with a professor on a research topic. Review the links for more research opportunities.