ABSTRACT. In manga, anime, and video games, there are representations of what are called energy waves. It is an expression in which a mass of light moves against a target point accompanied by luminescence, and is commonly used as a type of attack method. When anime were created using celluloid pictures, a photographic technique called transmitted light was used to achieve luminous expressions. The purpose of this research is to perform real-time rendering of transmitted light processing done during celluloid painting.

ABSTRACT. With the introduction of the metaverse, there is a growing need for a virtual space that can be shared by a large number of users for communication. In this paper, we propose an image-based distributed rendering architecture capable of rendering various types of models simultaneously, enabling multiple users to share a virtual space. To enhance the realism of this virtual environment, we introduce global illumination computation within this architecture, using screen-space technology and deferred shading. Finally, we present measurement results obtained from the implementation of our architecture on the grid computers of TSUBAME3.0.

ABSTRACT. We present a method to visualize a black hole scene at high speed by using a symplectic method with adaptive step size. Satoh et al. use the symplectic method to solve Hamilton’s canonical equation accurately with fixed step size. On the other hand, James et al. and Velasquez et al. use adaptive step size control, such as the Runge-Kutta Fehlberg method, to compute faster. Their approach is inferior to the symplectic method in terms of accuracy. Our approach incorporates the advantages of both methods. We compare the accuracy, computation time, and image quality to evaluate the usefulness of our method.

ABSTRACT. Volume rendering has been an important part of scientific data visualization in simulations of medical anatomy and analysis of internal physical quantities. Reducing the computational cost and time consumption of rendering complex volumes using this rendering method is one of key challenges for many data visualization researchers. Particle-based rendering (PBR) method has been proposed as one of the volume rendering methods. PBR is an efficient method, which enables to render semi-transparency for both of geometric and volumetric objects without visibility ordering. In this method, the objects can be rendered repeatedly based on stochastic process, and then the final image can be obtained by ensemble averaging the rendered images. The higher the number of repetitions, the higher the render quality. However, this can be a time-consuming process and a major cause of performance bottleneck, especially in the case of large-scale and complex dataset. To overcome these problems, we propose Multi-kernel denoising autoencoder (MkDAE) to denoise objects rendered with low repetitions to obtain high quality images as an alternative to rendering with high repetitions, reducing the time consumption.

ABSTRACT. Japanese swords are made through a repeated process of beating the steel, making cuts in it, and folding it in half. This process is called the fold-and-forging process. Kitaehada is a pattern created when a layer of steel produced by the fold-and-forging process appears on the surface. Before the fold-and-forging process, we represent a small steel pile as a height field. We also imitate the incising process by dividing the area of the height field in two. We have confirmed that different types of Kitaehada can be generated by changing the division direction in the iterations.

ABSTRACT. Motion capture is the primary means of acquiring movement. However, recording dangerous motions is burdensome for the performer. An alternative is physical simulation. Simply simulating a puppet that imitates a human being will result in unnatural bending of the joints. The range of motion of actual joints is complex and cannot be represented by simply setting a range for each axis. To solve this problem, we developed an improved physics simulation method that consists of passive restrictions using existing geometric joint ranges of motion and active restrictions that soften rotations that would exceed the ranges of motion in advance. We also implemented and validated this method in the Open Dynamics Engine.

ABSTRACT. We have performed a simulation of blood flow and pressure changes in the left ventricle and the aorta. In the previous simulation, pressure changes were almost the same as the real data; however, the simulated pressure in the aorta was a little bit lower than the real value after the valve closing because the aorta model was shorter than the real and the reflection flow from the peripheral vessel was not considered. This paper reports a new model that has a blood flow control mechanism, which accelerates the flow from the aorta to the peripherals and imitates the reflection. We have also visualized the stress distribution according to the opening and closing of the aortic valve.

ABSTRACT. One of the most common natural landscapes in our daily life is a landscape with trees. Tree models are an important research topic in the field of modeling in computer graphics. Crown shyness refers to the phenomenon in which tree branches grow without touching each other. In order to reproduce this phenomenon, this paper applies a space colonization algorithm, which makes it possible to realistically represent the beautiful appearance and flourishing of trees with crown shyness.

ABSTRACT. In this study, we propose a CG representation method of tire smoke during drifting based on physical simulations. The mechanism of how tire smoke is generated has not been fully elucidated. We calculated the generation and conduction of heat from tire rotation speed and ground friction, and constructed a model of tire smoke generation considering the frictional heat from the side surface when the vehicle is drifting. We confirmed the accuracy of the model by comparing the validity of temperature changes and the appearance with reality.

ABSTRACT. In computer animation, it is difficult to edit camera work. In this paper, we propose an intuitive method for creating a camera work by placing 3D printed figures to represent a scene and manipulating a virtual camera. As an example, we introduce a method using web cameras and AR markers. A character is positioned in a virtual space by shooting an AR marker attached to the bottom of a 3D printed figure from below with a web camera. Then, by shooting another AR marker with a different web camera, the position and direction of the virtual camera are acquired, and the camera in the virtual space is manipulated.