Moon-Ryul Jung

Hierarchical Retargetting of Fine Facial Motions

We present a novel technique for retargetting captured facial animation to new facial models. We use dense motion data that can express fine motions such as wrinkles. We use a normal mesh, which is a special multi‐resolution mesh, to represent source and target models. A normal mesh is defined by the base mesh and sequence of normal offsets from it. Our retargetting consists of two steps: base mesh and detail mesh retargetting. For base mesh retargetting, we use an example‐based technique to take advantage of the intuition of the user in specifying the similarity between source and target expressions. In detail mesh retargetting, the variations of normal offsets in the source mesh are hierarchically computed and transferred to the target mesh. Our retargetting scheme is able to produce robust and delicate results for unusual target models from highly detailed motion data.

Snaking across 3D meshes

We present a method of moving snake curves over 3D meshes. Our snake is made from the vertices of a triangular mesh, so it never leaves the 3D surface, eliminating any need for mapping on to ID and subsequent remapping on to the 3D mesh. Our snake finds features related to Gaussian curvature, or ridges and valleys. We move each vertex of the snake to one of its neighbor vertices on the mesh, while reducing the energy of the snake, which expresses distance from a feature, by means of a greedy optimization. This is fast enough to support real-time interaction. We handle changes of snake topology using 3D versions of methods originally developed for images. These can handle collisions within the snake, and hence split the snake and find multiple features. We demonstrate the performance of our algorithm on several example meshes, including human faces.

An Active Contour Approach to Extract Feature Regions from Triangular Meshes

We present a novel active contour-based two-pass approach to extract smooth feature regions from a triangular mesh. In the first pass, an active contour formulated in level-set surfaces is devised to extract feature regions with rough boundaries. In the second pass, the rough boundary curve is smoothed by minimizing internal energy, which is derived from its curvature. The separation of the extraction and smoothing process enables us to extract feature regions with smooth boundaries from a triangular mesh without users initial model. Furthermore, smooth feature curves can also be obtained by skeletonizing the smooth feature regions. We tested our algorithm on facial models and proved its excellence.

Local shape blending using coherent weighted regions

We present a novel local shape blending method that maps a sparse configuration of facial markers captured from an actor onto target models. The advantage of local shape blending methods is that, given a small set of key shapes for each local region, their combination can generate various facial expressions. However, they have the common problem that they use the pre-determined (fixed) regions and compute the combination of local key shapes for each region independently of each other. So, they have a risk of breaking natural correlations between the regions. We present a stochastic method of computing the regions and the blending weight vectors simultaneously. To do so, we formulate local shape blending as a problem of finding an optimal distribution of blending weight vectors of all control points in MAP–MRF framework.

Active contours with level-set for extracting feature curves from triangular meshes

In this paper, we present a novel algorithm that extracts feature curves from triangular mesh domains. It is an extension of the level-set formulation of active contour model in image space to triangular mesh domains. We assume that meshes handled by our method are smooth overall, and feature curves of meshes are thin regions rather than mathematical curves such as found in mechanical parts. We use a simple and robust scheme that assigns feature weights to the vertices of a mesh. We define the energy functional of the active contour over the domain of triangular mesh and derive a level-set evolution equation that finds feature regions. The feature regions are skeletonized and smoothed to form a set of smooth feature curves on the mesh.

Weighted local shape blending for facial motion retargetting

Realistic character animation requires elaborate rigging built on top of high quality 3D models. Sophisticated anatomically based rigs are often the choice of visual effect studios where life-like animation of CG characters is the primary objective. However, rigging a character with a muscular-skeletal system is very involving and time-consuming process, even for professionals. Although, there have been recent research efforts to automate either all or some parts of the rigging process, the complexity of anatomically based rigging nonetheless opens up new research challenges. We propose a new method to automate anatomically based rigging that transfers an existing rig of one character to another. The method is based on a data interpolation in the surface and volume domain, where various rigging elements can be transferred between different models. As it only requires a small number of corresponding input feature points, users can produce highly detailed rigs for a variety of desired character with ease. Copyright

Curved ray-casting for displacement mapping in the GPU

To achieve interactive speed, displacement mapping in the GPU is typically implemented in two steps: vertex shading/rasterization of the base surface and pixel shading. Pixel shading applies the height map relative to the image plane of the base surface, casts view rays to the height field through each pixel, finds the intersection point with the height field, and computes the color of that point. Here, the ray-casting process involves significant errors; The spatial relationship between the ray and the base surface is not preserved between the ray and the image plane of the base surface. The errors result in incorrect silhouettes. To address this problem, we curve the ray so that the spatial relationship between the (linear) ray and the base surface is preserved between the curved ray and the image plane of the base surface. This method reduces intersection errors, producing more satisfactory silhouettes, selfocclusions and shadows.

The Mail Art Box: An analog to digital postal service

This paper introduces an interactive media art work called the Mail Art Box that converts a handwritten message to a digital message to be transmitted to another person digitally. This paper explored the technology of taking a postcard picture in the embedded system and sending it via a wireless communication network. The work includes computer vision technology of recognizing the information of the postcard and extracting handwriting. The Mail Art Box was installed in a gallery SAI.

Three-dimensional graphic physically based simulator of rainbows together with the background scene

This paper presents a single scattering 3D graphics simulator of rainbows that includes the thickness of the rain shaft and the background scenery. The simulator is devised so that we can find a good configuration of the sun, the viewers, and the volume of water drops in a complicated geometric setting. The background-scene geometry and light-reflecting properties are modelled using 3D graphics tools. The simulator allows both the light reflected from the background surface and the light scattered by water drops to contribute to the final image by taking the depth to the background surface into account. The simulator generates an image of the rainbow by using the radiative transfer equation (RTE). We use ray optics to compute the average scattering cross section and the average phase function of particles that are the main parameters of the RTE. Depending on the density distribution of the water drops, the rainbow is perceived to be translucent, and the background scene is visible through the rainbow. We simulate other effects of the variation of the water-dropdensity and the location of the viewer, e.g., the visibility of the secondary rainbow, the brightness of the sky around the rainbow, the close-up view of the rainbow, and the full-circle rainbow. We explain these effects partly by computing the luminance contrasts of the primary and secondary bows against their local backgrounds.

An implicit active contour model for feature regions and lines

We present a level-set based implicit active contour method which can detect innermost homogeneous regions which are often considered feature regions or lines depending on the width of the regions. The curve evolution algorithm is derived from optimization of energy defined for the evolving curves. The energy has basically three terms: the first is the energy of the regions inside the curves, the second the energy of the bands inside the curves, and the third the energy of the bands outside the curves. If the band width is small, the total energy is minimized when the evolving curves lie at the boundaries of the innermost homogeneous regions, and the regions inside the curves are considered feature regions. Our method contrasts with the Chan-Vese model, which does not have the notion of innermost homogeneous regions but tries to find the curves such that the regions inside and outside them are both homogeneous. Our model approaches Chan-Vese model as the band width is increased, and is equivalent to Chan-Vese model when the band width is sufficiently large, so that all points inside/outside the curves lie within the bands inside/outside the curves, respectively.