Parallel ray tracing of complex scenes on multicomputers requires distribution of both computatio... more Parallel ray tracing of complex scenes on multicomputers requires distribution of both computations and scene data to the processors. This is carried out during preprocessing and usually consumes too much time and memory. In this paper, we present an efficient parallel subdivision algorithm to decompose a given scene into rectangular regions adaptively and map the resultant regions to the node processors of a multicomputer. The proposed algorithm uses efficient data structures to find out the splitting planes quickly. Furthermore the mapping of the regions and the objects to the node processors is being performed while parallel spatial subdivision proceeds. The proposed algorithm is implemented on an Intel's iPSC/2 hypercube multicomputer and promising results are obtained.
Soccer has always been one of the most popular videogame genres. When designing a soccer game, de... more Soccer has always been one of the most popular videogame genres. When designing a soccer game, designers tend to focus on the game field and game play due to the limited computational resources, and thus the modelling of virtual spectators is paid less attention. In this study we present a novel approach to the modeling of spectator behavior, which treats each spectator as a unique individual. We also propose an independent software layer for sport-based games that simply obtains the game status from the game engine via a simple messaging protocol and computes the spectator behavior accordingly. The result is returned to the game engine, to be used in the animation and rendering of the spectators. Additionally, we offer a customizable spectator knowledge base with well structured XML to minimize coding efforts, while generating individualized behavior. The employed AI is based on fuzzy inference. In order to overcome additional demand for computing realistic spectator behavior, we use GPU parallel computing with CUDA.
Özet: Üç boyutlu sanal dünyalar yüksek öğretimde yeni fırsatlar sunmaktadır. Dünyadaki birçok üni... more Özet: Üç boyutlu sanal dünyalar yüksek öğretimde yeni fırsatlar sunmaktadır. Dünyadaki birçok üniversite üç boyutlu sanal dünyalar arasında en yaygın olarak kullanılan Second Life platformunu eğitim amaçlı kullanmakta ve Second Life sanal kampüslerinde dersler vermekte, toplantılar, seminerler ve konferanslar düzenlenmektedir. Bu çalışmada, Second Life ODTÜ kampüsü tasarlanıp ve geliştirilmiştir. 2010 Sonbahar döneminde, ODTÜ Bilgisayar ve Öğretim Teknolojileri Eğitimi Bölümü üçüncü sınıf Özel Öğretim Yöntemleri dersi harmanlanmış olarak bu ortamda verilmiştir. Öğrenciler, grup aktivitelerinde yer almış ve öğretim tekniklerini sanal sınıflarda uygulamışlardır. Anahtar Sözcükler: Üç boyutlu sanal dünya, Second Life, Yüksek Öğretim Abstract: 3D virtual worlds provide new opportunities in higher education. Many universities in the world have been using Second Life, commonly used 3D virtual world, for educational purposes and they have been offering courses, conducting meetings, seminars and conferences in their Second Life Campuses. In this study, Second Life METU Campus has been designed and developed. During 2010 Fall semester, METU Computer Education and Instructional Technology Department Teaching Methods Course has been delivered in a blended way. Students participated in group activities and practiced the implementation of teaching techniques in virtual classrooms.
Massively multiplayer online games (MMOGs) have become highly popular in the last decade and now ... more Massively multiplayer online games (MMOGs) have become highly popular in the last decade and now attract millions of users from all over the world to play in an evolving virtual world concurrently over the Internet. The high popularity of MMOGs created a rapidly growing market and this highly dynamic market has forced the game developers to step up competitively. However, MMOG development is a challenging and expensive process. In this study, we have developed a network simulation tool which can be used to model and simulate typical MMOGs that have client-server architectures. Main objective is to provide a simulation environment to MMOG developers that could be used to test, analyze and verify various aspects of the MMOG network architecture. We have also implemented a graphical user interface which allows constructing the simulation model visually. We have demonstrated the use of simulation tool by experimental simulations.
PRESENCE: Teleoperators and Virtual …, Jan 1, 2010
With interest in virtual reality (VR) technologies, techniques, and devices growing at a quick pa... more With interest in virtual reality (VR) technologies, techniques, and devices growing at a quick pace, many researchers in areas such as psychology or cognitive neuroscience want to use VR. The software and VR systems available today do not support the skill sets or experimental requirements of this group of users. We describe a number of concerns and requirements that researchers
Three-dimensional computer animation can be very useful in presenting weather forecast on TV chan... more Three-dimensional computer animation can be very useful in presenting weather forecast on TV channels. However, generation of the animation for weather forecast using general-purpose tools might be difficult and time-consuming since this kind of animation requires special visual effects and various animatable parameters. Furthermore, the animation is to be generated by ordinary users with no training in animation. In this paper, we present a three-dimensional computer animation system specifically designed for presenting weather forecast. The system is designed and implemented using object-oriented techniques that are very helpful for computer animation and motion abstraction. The animation system also incorporates multiresolution modeling and paging of huge terrains for fast rendering.
Abstract. The aim of this study is to develop a methodology for validating driving simulators in ... more Abstract. The aim of this study is to develop a methodology for validating driving simulators in terms of simulator usability. The methodology was used to validate a truck simulator designed for training truck drivers about economic fuel consumption. The participants were eight truck drivers. Interview and observation methods were used to gather data. The results of the study showed that drivers did not have difficulty to recognize the parts of the driving simulator. Also, they stated that driving the simulator was easy. However, they said that ...
Parallel Computational Fluid Dynamics …, Jan 1, 2009
... processed in one render. Moreover, due to programmability at the fragment level, packing attr... more ... processed in one render. Moreover, due to programmability at the fragment level, packing attributes into a single RGBA texel at the fragment level decreases the number of rendering passes by and Veysi Isler 385 DOI: 10.1007/978-3 ...
We propose a prefetching method to improve smoothness of client side rendering in distributed vir... more We propose a prefetching method to improve smoothness of client side rendering in distributed virtual urban environments. In order to find optimum prefetching policy, this method reduces the prefetching problem to a graph partitioning problem. The scene is represented by a graph where the vertices denote view-cells and the edges denote the adjacencies among the cells. We utilize the Simulated Annealing algorithm to find an approximation for the optimum partitioning. The proposed method aims to minimize view-cell transition cost by utilizing statistical data obtained from clients. Multi-agent client simulator has been implemented for testing the efficiency gain and detecting characteristic information about clients. Our algorithm iteratively refines the prefetching policy by using client tendancy data obtained from the simulator. Experimental results confirmed significant improvements on client rendering of virtual urban environments.
Although volume rendering is a powerful computer graphics technique for visualizing data represen... more Although volume rendering is a powerful computer graphics technique for visualizing data represented at three spatial dimensions, it is computationally intensive. Parallel Computing represents the most promising solutions for computationally expensive problems. This paper aims to introduce some new concepts and proposes a new method for parallel volume rendering on MIMD architectures. The method described uses dynamic load balancing techniques to achieve higher efficiency and especially focuses on efficient progressive redistribution of the data during the consecutive rendering of animation frames by taking view coherence into consideration. The method is named as intelligent because most of the necessary actions for rendering are automatically handled by decisions locally made on each processor. Almost all global decisions such as data redistribution scheme and synchronization are replaced with local decisions. The method also incorporates the overlapping of the interprocessor messaging times with useful computation times.
Proceedings of the 9th Conference on …, Jan 1, 2000
The objective of this study is to develop an intelligent action algorithm for virtual human agent... more The objective of this study is to develop an intelligent action algorithm for virtual human agents on three-dimensional large terrains to accomplish a specified mission by group communication and coordination. The area contains natural and build-in entities such as trees, rocks, rivers, roads, houses, bridges, etc. Our platoons that are represented by virtual human agents enter a specific area to perform a specified mission, which may be to attack, escape or just pass through a selected tactical area. The area contains static and/or moving platforms such as jeeps, planes, helicopters, commandos, and etc. The goal of the agents is to complete their mission in a group or by being divided into groups of two or more without being detected or caught by a platform that carries different kinds of sensors (Day TV, Infra-Red, SAR). The output views of the platform sensors are observed by the user at tactical command center in order to make the detection process realistic. Agents may follow rivers, go through the forests, and hide behind trees, run, or even crawl in order not to be seen. When any of the agents are detected and identified, they try to escape or hide to complete their mission until they are caught or terminated.
In the recent years Graphics Processing Units (GPU) have evolved into general purpose programmabl... more In the recent years Graphics Processing Units (GPU) have evolved into general purpose programmable streaming parallel processors. This evolution makes it possible to implement high quality photo realistic rendering techniques on graphics processors. There has been a few studies to ...
Parallel ray tracing of complex scenes on multicomputers requires distribution of both computatio... more Parallel ray tracing of complex scenes on multicomputers requires distribution of both computations and scene data to the processors. This is carried out during preprocessing and usually consumes too much time and memory. In this paper, we present an efficient parallel subdivision algorithm to decompose a given scene into rectangular regions adaptively and map the resultant regions to the node processors of a multicomputer. The proposed algorithm uses efficient data structures to find out the splitting planes quickly. Furthermore the mapping of the regions and the objects to the node processors is being performed while parallel spatial subdivision proceeds. The proposed algorithm is implemented on an Intel's iPSC/2 hypercube multicomputer and promising results are obtained.
An algorithm for locating different types of sensors in a 3D virtual environment to achieve a des... more An algorithm for locating different types of sensors in a 3D virtual environment to achieve a desired coverage of that environment while satisfying some constraints is being developed. In this context, the keyword 'sensor' is used to identify a FLIR (Forward Looking Infra-Red) camera. It can be extended to include Day-TV cameras, night vision goggles and radars. In this problem, the sensor system is composed of two major types of sensor platforms, namely moving and static platforms. The main goal of this system is to maintain security of the overall or indicated regions of environment. While achieving this goal, the system should also satisfy constraints imposed on the system. Maintaining stealth (remaining hidden), constraining route and placement locations are examples of these constraints. As a result of the optimization of the sensor system, the optimal routes and viewing directions for the moving sensors, the optimal locations and viewing directions for the static sensors will be found.
We propose a method for pre-fetching scene data to improve smoothness of client side rendering in... more We propose a method for pre-fetching scene data to improve smoothness of client side rendering in distributed urban environments. The scene is represented as a graph where the nodes are the view-cells and the edges denote the adjacencies among the cells. Prefetching is ...
Parallel ray tracing of complex scenes on multicomputers requires distribution of both computatio... more Parallel ray tracing of complex scenes on multicomputers requires distribution of both computations and scene data to the processors. This is carried out during preprocessing and usually consumes too much time and memory. In this paper, we present an efficient parallel subdivision algorithm to decompose a given scene into rectangular regions adaptively and map the resultant regions to the node processors of a multicomputer. The proposed algorithm uses efficient data structures to find out the splitting planes quickly. Furthermore the mapping of the regions and the objects to the node processors is being performed while parallel spatial subdivision proceeds. The proposed algorithm is implemented on an Intel's iPSC/2 hypercube multicomputer and promising results are obtained.
Soccer has always been one of the most popular videogame genres. When designing a soccer game, de... more Soccer has always been one of the most popular videogame genres. When designing a soccer game, designers tend to focus on the game field and game play due to the limited computational resources, and thus the modelling of virtual spectators is paid less attention. In this study we present a novel approach to the modeling of spectator behavior, which treats each spectator as a unique individual. We also propose an independent software layer for sport-based games that simply obtains the game status from the game engine via a simple messaging protocol and computes the spectator behavior accordingly. The result is returned to the game engine, to be used in the animation and rendering of the spectators. Additionally, we offer a customizable spectator knowledge base with well structured XML to minimize coding efforts, while generating individualized behavior. The employed AI is based on fuzzy inference. In order to overcome additional demand for computing realistic spectator behavior, we use GPU parallel computing with CUDA.
Özet: Üç boyutlu sanal dünyalar yüksek öğretimde yeni fırsatlar sunmaktadır. Dünyadaki birçok üni... more Özet: Üç boyutlu sanal dünyalar yüksek öğretimde yeni fırsatlar sunmaktadır. Dünyadaki birçok üniversite üç boyutlu sanal dünyalar arasında en yaygın olarak kullanılan Second Life platformunu eğitim amaçlı kullanmakta ve Second Life sanal kampüslerinde dersler vermekte, toplantılar, seminerler ve konferanslar düzenlenmektedir. Bu çalışmada, Second Life ODTÜ kampüsü tasarlanıp ve geliştirilmiştir. 2010 Sonbahar döneminde, ODTÜ Bilgisayar ve Öğretim Teknolojileri Eğitimi Bölümü üçüncü sınıf Özel Öğretim Yöntemleri dersi harmanlanmış olarak bu ortamda verilmiştir. Öğrenciler, grup aktivitelerinde yer almış ve öğretim tekniklerini sanal sınıflarda uygulamışlardır. Anahtar Sözcükler: Üç boyutlu sanal dünya, Second Life, Yüksek Öğretim Abstract: 3D virtual worlds provide new opportunities in higher education. Many universities in the world have been using Second Life, commonly used 3D virtual world, for educational purposes and they have been offering courses, conducting meetings, seminars and conferences in their Second Life Campuses. In this study, Second Life METU Campus has been designed and developed. During 2010 Fall semester, METU Computer Education and Instructional Technology Department Teaching Methods Course has been delivered in a blended way. Students participated in group activities and practiced the implementation of teaching techniques in virtual classrooms.
Massively multiplayer online games (MMOGs) have become highly popular in the last decade and now ... more Massively multiplayer online games (MMOGs) have become highly popular in the last decade and now attract millions of users from all over the world to play in an evolving virtual world concurrently over the Internet. The high popularity of MMOGs created a rapidly growing market and this highly dynamic market has forced the game developers to step up competitively. However, MMOG development is a challenging and expensive process. In this study, we have developed a network simulation tool which can be used to model and simulate typical MMOGs that have client-server architectures. Main objective is to provide a simulation environment to MMOG developers that could be used to test, analyze and verify various aspects of the MMOG network architecture. We have also implemented a graphical user interface which allows constructing the simulation model visually. We have demonstrated the use of simulation tool by experimental simulations.
PRESENCE: Teleoperators and Virtual …, Jan 1, 2010
With interest in virtual reality (VR) technologies, techniques, and devices growing at a quick pa... more With interest in virtual reality (VR) technologies, techniques, and devices growing at a quick pace, many researchers in areas such as psychology or cognitive neuroscience want to use VR. The software and VR systems available today do not support the skill sets or experimental requirements of this group of users. We describe a number of concerns and requirements that researchers
Three-dimensional computer animation can be very useful in presenting weather forecast on TV chan... more Three-dimensional computer animation can be very useful in presenting weather forecast on TV channels. However, generation of the animation for weather forecast using general-purpose tools might be difficult and time-consuming since this kind of animation requires special visual effects and various animatable parameters. Furthermore, the animation is to be generated by ordinary users with no training in animation. In this paper, we present a three-dimensional computer animation system specifically designed for presenting weather forecast. The system is designed and implemented using object-oriented techniques that are very helpful for computer animation and motion abstraction. The animation system also incorporates multiresolution modeling and paging of huge terrains for fast rendering.
Abstract. The aim of this study is to develop a methodology for validating driving simulators in ... more Abstract. The aim of this study is to develop a methodology for validating driving simulators in terms of simulator usability. The methodology was used to validate a truck simulator designed for training truck drivers about economic fuel consumption. The participants were eight truck drivers. Interview and observation methods were used to gather data. The results of the study showed that drivers did not have difficulty to recognize the parts of the driving simulator. Also, they stated that driving the simulator was easy. However, they said that ...
Parallel Computational Fluid Dynamics …, Jan 1, 2009
... processed in one render. Moreover, due to programmability at the fragment level, packing attr... more ... processed in one render. Moreover, due to programmability at the fragment level, packing attributes into a single RGBA texel at the fragment level decreases the number of rendering passes by and Veysi Isler 385 DOI: 10.1007/978-3 ...
We propose a prefetching method to improve smoothness of client side rendering in distributed vir... more We propose a prefetching method to improve smoothness of client side rendering in distributed virtual urban environments. In order to find optimum prefetching policy, this method reduces the prefetching problem to a graph partitioning problem. The scene is represented by a graph where the vertices denote view-cells and the edges denote the adjacencies among the cells. We utilize the Simulated Annealing algorithm to find an approximation for the optimum partitioning. The proposed method aims to minimize view-cell transition cost by utilizing statistical data obtained from clients. Multi-agent client simulator has been implemented for testing the efficiency gain and detecting characteristic information about clients. Our algorithm iteratively refines the prefetching policy by using client tendancy data obtained from the simulator. Experimental results confirmed significant improvements on client rendering of virtual urban environments.
Although volume rendering is a powerful computer graphics technique for visualizing data represen... more Although volume rendering is a powerful computer graphics technique for visualizing data represented at three spatial dimensions, it is computationally intensive. Parallel Computing represents the most promising solutions for computationally expensive problems. This paper aims to introduce some new concepts and proposes a new method for parallel volume rendering on MIMD architectures. The method described uses dynamic load balancing techniques to achieve higher efficiency and especially focuses on efficient progressive redistribution of the data during the consecutive rendering of animation frames by taking view coherence into consideration. The method is named as intelligent because most of the necessary actions for rendering are automatically handled by decisions locally made on each processor. Almost all global decisions such as data redistribution scheme and synchronization are replaced with local decisions. The method also incorporates the overlapping of the interprocessor messaging times with useful computation times.
Proceedings of the 9th Conference on …, Jan 1, 2000
The objective of this study is to develop an intelligent action algorithm for virtual human agent... more The objective of this study is to develop an intelligent action algorithm for virtual human agents on three-dimensional large terrains to accomplish a specified mission by group communication and coordination. The area contains natural and build-in entities such as trees, rocks, rivers, roads, houses, bridges, etc. Our platoons that are represented by virtual human agents enter a specific area to perform a specified mission, which may be to attack, escape or just pass through a selected tactical area. The area contains static and/or moving platforms such as jeeps, planes, helicopters, commandos, and etc. The goal of the agents is to complete their mission in a group or by being divided into groups of two or more without being detected or caught by a platform that carries different kinds of sensors (Day TV, Infra-Red, SAR). The output views of the platform sensors are observed by the user at tactical command center in order to make the detection process realistic. Agents may follow rivers, go through the forests, and hide behind trees, run, or even crawl in order not to be seen. When any of the agents are detected and identified, they try to escape or hide to complete their mission until they are caught or terminated.
In the recent years Graphics Processing Units (GPU) have evolved into general purpose programmabl... more In the recent years Graphics Processing Units (GPU) have evolved into general purpose programmable streaming parallel processors. This evolution makes it possible to implement high quality photo realistic rendering techniques on graphics processors. There has been a few studies to ...
Parallel ray tracing of complex scenes on multicomputers requires distribution of both computatio... more Parallel ray tracing of complex scenes on multicomputers requires distribution of both computations and scene data to the processors. This is carried out during preprocessing and usually consumes too much time and memory. In this paper, we present an efficient parallel subdivision algorithm to decompose a given scene into rectangular regions adaptively and map the resultant regions to the node processors of a multicomputer. The proposed algorithm uses efficient data structures to find out the splitting planes quickly. Furthermore the mapping of the regions and the objects to the node processors is being performed while parallel spatial subdivision proceeds. The proposed algorithm is implemented on an Intel's iPSC/2 hypercube multicomputer and promising results are obtained.
An algorithm for locating different types of sensors in a 3D virtual environment to achieve a des... more An algorithm for locating different types of sensors in a 3D virtual environment to achieve a desired coverage of that environment while satisfying some constraints is being developed. In this context, the keyword 'sensor' is used to identify a FLIR (Forward Looking Infra-Red) camera. It can be extended to include Day-TV cameras, night vision goggles and radars. In this problem, the sensor system is composed of two major types of sensor platforms, namely moving and static platforms. The main goal of this system is to maintain security of the overall or indicated regions of environment. While achieving this goal, the system should also satisfy constraints imposed on the system. Maintaining stealth (remaining hidden), constraining route and placement locations are examples of these constraints. As a result of the optimization of the sensor system, the optimal routes and viewing directions for the moving sensors, the optimal locations and viewing directions for the static sensors will be found.
We propose a method for pre-fetching scene data to improve smoothness of client side rendering in... more We propose a method for pre-fetching scene data to improve smoothness of client side rendering in distributed urban environments. The scene is represented as a graph where the nodes are the view-cells and the edges denote the adjacencies among the cells. Prefetching is ...
Uploads
Papers by Veysi Isler