A growing body of literature has emerged that demonstrates the potential of neurogaming platforms for inter- facing with well-known video games. With the recent convergence of advances in consumer electronics, ubiquitous computing, and wearable sensor technologies real-time monitoring of neurocognitive and affective states can be studied in an objective manner. Whilst establishing the optimal relation among frequency bands, task engagement, and arousal states is a goal of neurogaming, a standardized method has yet to be established. Herein we aimed to test classifiers within the same context, group of participants, feature extraction methods, and protocol. Given the emphasis upon neurogaming, a commercial-grade electroencephalographic (EEG; Emotiv EPOC) headset was used to collect signals from 30 participants. The EEG data was then filtered to get separate frequency bands to train cognitive-affective classifiers with three classification techniques: Support Vector Machines (SVM), Naive Bayes (NB), and k-Nearest Neighbors (kNN). Results revealed that the NB classifier was the most robust classifier for identifying negative (e.g., character death) game-based events. The identification of general gameplay events is best identified using kNN and the Beta band. Results from this study suggest that a combination of classifiers is preferable over selection of a single classifier.

While the 15-puzzle has a long and interesting history dating back to the 1870s, it still continues to appear as apps on mobile devices and as minigames inside larger video games. We demonstrate a method for solving the 15-puzzle using only 4.7 MB of tables that on a million random instances was able to find solutions of 65.21 moves on average and 95 moves in the worst case in under a tenth of a millisecond per solution on current desktop computing hardware. These numbers compare favorably to the worst case upper bound of 80 moves and to the greedy algorithm published in 1995, which required 118 moves on average and 195 moves in the worst case.

We present a framework for emulating the deformation and consumption of polygonal models under combustion while generating procedural fire. Our focus is on achieving the best visual effects possible while maximizing computation speed so that the processing power is available for other tasks in video games. We have implemented and tested our method on a relatively modest GPU using CUDA. Our experiments suggest that our method gives a believable rendering of the effects of fire while using only a small fraction of CPU and GPU resources.

Procedural terrain generation is used to create landforms for applications such as computer games and flight simulators. While most of the existing work has concentrated on algorithms that generate terrain without input from the user, we explore a more controllable system that uses intelligent agents to generate terrain elevation heightmaps according to designer-defined constraints. This allows the designer to create procedural terrain that has specific properties.

Computer science courses have increasingly made use of visualization tools to illustrate common algorithms. This paper describes Algorithm Explorer, an educational tool designed for use by instructors and students to examine algorithms in a rich environment composed of 3D data representations, 3D audio cues, and easy-to-use controls. Instructors and students can easily add calls to Algorithm Explorer's C++ API to their programs to quickly develop engaging visualizations, and every detail of the scene can be customized as the developer desires.

All in-text\treferences\tunderlined\tin\tblue\tare\tlinked\tto\tpublications\ton\tResearchGate, letting you\taccess\tand\tread\tthem\timmediately.

No abstract available.

Game programming classes have been offered at the University of North Texas continuously since 1993. The classes are project based, and feature collaborative coursework with art majors in UNT's School of Visual Arts. We discuss the design that enables them to simultaneously provide both training for students intending employment in the game industry, and a capstone experience for general computer science undergraduates.

The maze routing problem is to find an optimal path between a given pair of cells on a grid plane. Lee's algorithm and its variants, probably the most widely used maze routing method, fails to work in the 4-geometry of the grid plane. Our algorithm solves this problem by using a suitable data structure for uniform wave propagation in the 4-geometry, 8-geometry, etc. The algorithm guarantees finding an optimal path if it exists and has linear time and space complexities. Next, to solve the obstacle-avoiding rectilinear and 4-geometry Steiner tree problems, a heuristic algorithm is presented. The algorithm utilizes a cost accumulation scheme based on the maze router to determine the Torricelli vertices (points) for improving the quality of multiterminal nets. Our experimental results show that the algorithm works well in practice. Furthermore, using the 4-geometry router, path lengths can be significantly reduced up to 12% compared to those in the rectilinear router.

It is shown that in a standard discrete neural network model with small fan-in, tolerance to random malicious faults can be achieved with a log-linear increase in the number of neurons and a constant factor increase in parallel time, provided fan-in can increase arbitrarily. A similar result is obtained for a nonstandard but closely related model with no restriction on fan-in.

The time required by a variant of the PRAM (a parallel machine model which consists of sequential processors which communicate by reading and writing into a common shared memory) to compute a certain class of functions called critical functions (which include the Boolean OR of n bits) is studied. Simultaneous reads from individual cells of the shared memory are permitted, but simultaneous writes are not. It is shown that any PRAM which computes a critical function must take at least $0.5\log n - O(1)$ steps, and that there exists a critical function which can be computed in $0.57\log n - O(1)$ steps. These bounds represent an improvement in the constant factor over those previously known.

It is demonstrated that there is no nine-input sorting network of depth six. The proof was obtained by executing on a supercomputer a branch-and-bound algorithm which constructs and tests a critical subset of all possible candidates. Such proofs can be classified as experimental science, rather than mathematics. In keeping with the paradigms of experimental science, a high-level description of the experiment and analysis of the result are given.

Nondeterministic analogues of the well-known language classes NC and SC called NNC and NSC, respectively, are investigated. NC is the class of languages that can be accepted by small, fast parallel computers; SC is the class of languages that can be recognized by a deterministic Turing machine in polynomial time and polylog tape-head reversals. Adding nondeterminism to SC leaves it in the domain of parallel computation since NSC contained in POLYLOGSPACE. That is, NSC is a subset of the class of languages computable by fast parallel computers. Adding nondeterminism to NC appears to make it much more powerful since NNC=NP. It is clear that NSC contained in NNC, and probable that NSC contained in/implied by NNC. Further evidence for this conjecture is provided by showing that NSC is precisely the class of languages recognizable in simultaneous polynomial time and polylog reversals by a nondeterministic Turing machine with a read-only input tape and a single read-write work tape; it is known that NNC is similar, but is recognizable by a Turing machine with two read-write tapes.>