This paper describes a method for rapidly authoring multiplayer online games. The game author specifies state machines that define the behavior of game characters. The results are translated into a JavaScript code. The entire application is supported by Node.js at the server side and by modern Web browsers at the client side. Utilization of the WebSocket feature allows the communication between the Web browsers and the web server to be processed efficiently, and the player's interaction with the game interface is smooth.

A remote view system is a system for the delivery and presentation of information such as video and audio to users in remote locations. Methods for the construction of virtual environments based on video information have been applied to remote view technology. When a system that supports remote viewing is applied to situations such as viewing scenery or engaging in collaborative work, it is useful to provide users with a function that enables them to walk through a virtual environment. Although walkthroughs are possible in remote view systems that use pre-prepared images to construct virtual environments, these systems are lacking in terms of their real-time performance. In this study, we built a virtual walk-through environment that allows remote viewing based on images sent in real time from an omni directional camera. By using an improved map projection method, the system produces less image distortion than the conventional projection method. We have confirmed the walk-through capabilities of this system, evaluated the performance of improved map projection method, and subjected it to user trials.

In recent times, the web-based applications has grown rapidly and there has been an explosive increase in the number of users. Hence, it is necessary to develop technique to deal with several users. In this study, we particularly focus on web-based multiplayer online role-playing games (MORPGs), which have a large number of users and deal with high frequency of user requests. In our previous work, we introduced a method to deal with a large number of users using multiple web servers and a dynamic data allocation method to deal with dynamic change in the user request. Using our method, we achieved a user capacity of 320 for the web-based MORPG system as opposed to 200 user capacity of the single server MORPG system. However, we found some cases in which our dynamic data allocation role was not effective. In this paper, we propose three rules for our dynamic data allocation method to address various situations. We evaluate these rules using four web servers and investigate their properties and effectiveness.

Since the departure of users has a large effect on the integrity of shared files in a P2P system, there is a strong need for methods that can mitigate the effects of user departure. In this paper, introducing a graph to represent the file-sharing in a P2P system, we model the participation and departure processes of users with a renewal process to determine the node departure time. The fact that the departure interval of nodes in the introduced connection graph can be expressed in terms of residual lifetimes is used to derive the expected values of node departure time for several lifetime distributions of nodes. It is also shown that the Laplace transform of the number of departure nodes can be expressed in terms of the node lifetime distribution function.

Communication latency with the lockstep protocol is minimized under the assumptions that (1) time-stamp servers are available near each player that issue serially numbered time stamps and that (2) there is no communication break down between the player and the time-stamp server. Those assumptions are relatively moderate since the time-stamping service is becoming the fundamental infrastructure of the cyberworlds. Each player sends its own action to the other player and also sends its hash to the nearest time-stamp server. The time-stamp server sends back to the player the signed hash with time and a serial number involved. The signature is an undeniable evidence of the action. The actions are checked if they are compatible with the hashes and the signed hashes are checked if they have the correct time and the serial numbers are contiguous. This verification can be done as a batch after the game is finished. While the original lockstep protocol suffers from the latency as much as the round-trip time of packets and the other existing improved protocols alleviate the latency at the cost of lower security, the proposed protocol has the latency as little as the packet travelling time from one player to another and has the same security as the original under the assumption that we trust the time-stamp servers.

Over the last few years a great deal of interest has focused on a major paradigm shift in World Wide Web-based services referred to as Web 2.0. In marked contrast to the earlier Web, in Web 2.0 sites become sources of information and functionality that enable users to create new content of their own. Consequently, users are now looking for more versatile browsers that will let them edit and display content based on their own preferences. This motivated us to develop a state diagram-based Web browser programming scheme that supports the close interaction between the end-user and Web content. Using state diagrams to represent browser behavior is easier for end-users with little or no programming experience to grasp than text-based programming systems.

In the virtual ball game, such as soccer, football, basketball and hockey etc. in a virtual space, "pass" plays have a significant meaning for players to build up their games. Goodness of a passes course and its velocity is continuously varied, both spatially and temporarily by all motions of ally player avatars and opponent player avatars attending in the game. Because of this, a significant pass between the avatars may not easily be able to be analyzed and generated by the computer systems in real-time. In this paper, we propose a new algorithm that can realize this computation in real-time. Concretely, a real-time detection of a "space" or an "open space" for every avatar by using the idea of "AtoZ (Allocated Topographical Zone)" that Kawano et al., have already proposed. Based on this idea, we compute the reachable pass vector to any avatar in the game. Moreover, the space where a player can receive a pass given by the pass vector, prior to any other player, is defined as an "exclusively reachable pass zone" for the receiver, and is able to be computed by the algorithm.

In this paper, we studied the relationship between the accuracy of the dead reckoning data and the update interval in a distributed virtual environment (DVE). We have already shown that numerical analysis can be applied to the dead reckoning data between frames using parametrics calculated from the data over the last several frames based on polynomial models. Based on the above polynomial models, we proposed the new dead reckoning method to extrapolate the attribute data which arrives at discrete time period. We showed that theoretical models which approximate the statistical error of dead reckoning data can be formulated based on parameters such as the update interval and changes in the data. Theoretical models showed that the average error of proposed method is less than that of current method. Finally, we evaluate the validity of the proposed method with the comparison of current method by conducting experiments with the pen motion of a series of letters written by human. We confirmed that the proposed method can decrease the extrapolation error in comparison with current method in average

A system of vision-based mixed reality is developed for implementing outdoor gaming on a mobile device having a single camera and a slow CPU. To help the poor device, it is assumed that sufficiently many marker objects of the same size are scattered in the real world so that relative translation and rotation of the camera can be easily calculated from their position and size in the camera image. This system can be applied to outdoor games of which scenario allows for existence of some items that the markers can mimic

In this paper, the authors aim to analyze the error of dead reckoned data generated from received data in a discrete temporal axis in a distributed virtual environment (DVE). That is, compared with the data received in continuous time, data acquired in discrete time has a certain degradation or uncertainty of information. Our way of analysis is to introduce a mathematical model of this degradation with regard to the metrics of the temporal interval. We introduced polynomial models for dead reckoning method between frames using parametrics calculated from the data over the last several frames. By employing the method of error analysis of a numerical analysis to the above polynomial models, we formulate theoretical models which approximate the statistical error of dead reckoned data based on parameters such as the update interval and changes in the data. This study enables the discussion on the optimality of the update interval in a DVE using dead reckoning. Finally, we evaluate the adaptability of the theoretical model by conducting simulation experiments generated by pen motion of writing string of letters by human. As a result, we confirm that the proposed theoretical model closely approximates the average error in the simulation

We have developed a system called "Web-Com "for Web-based education: the Web browser provides a multi-layer drawable canvas on which users can draw annotations on Web pages. Annotations can be shared with other users in real-time and non-real-time for use of synchronous and asynchronous learning. This paper mainly discusses the asynchronous function including processing of annotation by recorded voice. That is, it can replay the process of annotations including the process of voice recording, which enables asynchronous learning. In our previous work, we conducted a subject experiment to evaluate Web-Com's synchronous functions. The experimental results showed that the learners could learn fairly interactively with an instructor in a Web-based class. In this paper, we conducted a subject experiment in an actual class, to evaluate its asynchronous functions and validity of asynchronous learning. The experimental results indicate that Web-Com has advantages for the learners to understand given lectures

This paper studies the way to improve consistency of states in a ball-game typed distributed virtual environment (DVE) with lag, in peer-to-peer (P2P) architecture. That is, we are studying how to reduce in real-time the difference of states between the participating terminals in a virtual ball game caused by transmission lag or update interval. We are also studying how to control shared objects in real-time in a server-less network architecture. Specifically, a priority field called Allocated Topographical Zone (AtoZ) is used in P2P for DVE. In a critical case, defined as inconsistent phenomena between the peers, caused by the network latency, a stricter ownership determination algorithm, called dead zone is introduced. Concession parameter is used to reduce the possibility of such a case. By using these protocols in combination, a robust and effective scheme is achieved for a virtual ball game and these protocols ware extended for multi-player game. As an example of the application, a realtime networked doubles air-hockey is implemented for evaluation of the influence of these protocols on interactivity and on consistency

Recently, various types of personal authentication methods have been studied to confirm a persons' identity in the on-line system. However, the alphanumeric characters and authentication methods based on passwords and PINs (knowledge-factor authenticators) hold severe problems, and still must rely on the limitation of human's capacity of recollection. Recently, "graphical password", using visual information for authentication methods that utilize photographic image for authentication, has been proposed. However this method holds some problem on security. In this paper, we propose "COMPASS" (community portrait authentication system) that uses a portrait as the authentication image, to solve the issues of the previous graphical passwords. Moreover, COMPASS has the idea of the "community authentication" that a member of a certain community consisting of a party of human can be authenticated. The validity of the proposed method is confirmed by a set of subject experiments

The authors introduce a newly developed Web-based education system. The developed system supports both the synchronous learning and the asynchronous learning. The system consists of the interactive Web browser called "Web-Com" and the voice server. The Web-Com provides with the multilayer canvas and the user can draw annotations on the layer. Each layer can be shared with other user in real-time, via the network, to be utilized for the synchronous learning. In conjunction with the voice server, the Web-Com can support the communication by human voice. The Web-Com also supports the function that can replay the process of annotations in their order, which enables the asynchronous learning. Finally, a subject experiment is performed to examine the validity of the system. The result of the experiment shows that the user can learn interactively in the Web-based class by using the Web-Com.

This paper aims to study how to gain consistency of the state in a DVE (distributed virtual environment) with lag, under the condition of peer-to-peer architecture. That is, we study how to reduce in real-time the difference of the DVE state between peers caused by transmission lag or update interval, and how to control the shared object in the online field game. Specifically, the priority field called allocated topographical zone (AtoZ) and mutual prediction protocol is used in peer-to-peer DVE. Moreover considering the critical case (the inconsistent phenomena) caused by the network latency between peers, stricter ownership determination algorithm called the dead zone is introduced. Using these three protocols in combination, a robust and effective scheme is realized. As an example of the application using the protocols, a real-time networked air-hockey is implemented for evaluation of the influence on interactivity and consistency.

When a musical session is performed via networks, it is necessary to interact in real-time. There exists, however, the problem of delay between the nodes, which causes musical notes become an impediment. To overcome this problem, we had proposed a new protocol for musical session called mutual anticipated session (M.A.S.), which is a type of ensemble that controls the appropriate timing of the sounds. In the M.A.S, one player's performance precedes the other players', so we call this performance "precedent musical performance", and we call the time lapse between the players' performance as "precedent time." In the current M.A.S, it is assumed that the tempo during the performance is constant. In a case, however, where players want to perform more expressively or more emotionally, they may change the tempo. Thus, in this paper, we add a function that accommodates the change of the tempo and evaluate the function experimentally.

The paper introduces the actualization of a new communication channel under a pseudo-real-time phenomenon, in an augmented distributed virtual environment (ADVE) framework through the implementation of a prediction feedback loop (PFL), upon the existence of the unavoidable lag derived from the natural and artificial temporal factors. By implementing PFL, the ADVE framework is enriched with predictive characteristic where it is able to predict and display information of a remotely controlled avatar (virtual body) in the ADVE. Based on the predictive results, interactions between a locally controlled avatar and a remotely controlled avatar in the ADVE can be carried out smoothly, effectively and strategically under an illusory real time phenomenon without influence from the existence of lag. With this pseudo-real-time channel, we would like to bring forth a new era of effective man-to-man virtual communication over the network without inefficiency caused by the unavoidable lag, whether it is solvable or unsolvable.

This paper proposes a new three-dimensional pointing device considering user friendliness and handiness. The proposed method utilizes five degrees of freedom via the medium of nonverbal voice of human. That is, the spatial direction of the sound source, the type of the voice phoneme and the "tone" of the voice phoneme. The input voice is analyzed regarding the above factors and then taking proper effects as previously defined for human interface. In this paper the estimated spatial direction is used for three-dimensional movement for the virtual object as three degrees of freedom. Both of the type and the tone of the voice phoneme are used for remaining two degrees of freedom. Since vocalization of nonverbal human voice is an everyday task, and the intonation of the voice can be quite easily and intentionally controlled by human vocal ability, the proposed scheme is a new three-dimensional spatial interaction medium. In this sense, this paper realizes a cost-effective and handy nonverbal interface scheme without any artificial wearing materials which might give a physical and psychological fatigue.