WSCG 2014: Full Papers Proceedings
Permanent URI for this collection
Browse
Recent Submissions
Item Simulating Human Vision and Correcting Visual Aberrations with Computational Light Field Displays(Václav Skala - UNION Agency, 2014) Barsky, Brian A.; Skala, VáclavThis talk will present research on simulating human vision and on correcting visual aberrations with computational light field displays . The simulation is not an abstract model but incorporates real measurements of a particular individual’s entire optical system. Using these measurements, synthetics images are generated. This process modifies input images to simulate the appearance of the scene for the individual. Recent work on vision-correcting displays will also be briefly introduced. Given the measurements of the optical aberrations of a user’s eye, a vision correcting display will present a transformed image that when viewed by this individual will appear in sharp focus. This could impact computer monitors, laptops, tablets, and mobile phones. Vision correction could be provided in some cases where spectacles are ineffective.
Item Flow maps - Benefits, Problems, Future Research(Václav Skala - UNION Agency, 2014) Weinkauf, Tino; Skala, VáclavThe flow map has become a standard tool for the analysis and visualization of unsteady flows. In simple terms, it maps the start point of a particle integration to its end point. Flow maps are used to compute Finite Time Lyapunov Exponents (FTLE), Streak Line Vector Fields, or to speed up other methods in flow visualization. However, they are very costly in terms of both computation time and storage. In this talk, I will give an overview of the latest developments in flow visualization, review the theoretical and practical benefits of flow maps, discuss issues of accuracy and complexity, and pose open questions for future research in this area.Item Performing High-Dimensional Filtering in Low-Dimensional Spaces(Václav Skala - UNION Agency, 2014) Oliveira, Manuel M.; Skala, VáclavHigh-dimensional filtering is a key component for many graphics, image, and video processing applications. Edge-preserving filters (an important class of high-dimensional ones), for instance, are essential for tasks like global-illumination filtering, tone mapping, denoising, detail enhancement, and non-photorealistic effects, among many others. Edge-preserving filtering can be implemented as a convolution with a spatially-varying kernel in image space, or with a spatially-invariant kernel in high-dimensional space. Performing the operation either way is computationally expensive, preventing its use in interactive and real-time scenarios. The talk will present two recent techniques we have developed for efficiently performing edgeaware filtering. The first one is based on a domain transform that allows highdimensional geodesic filtering to be performed in linear time as a sequence of 1-D filtering steps using a spatially-invariant kernel. The second technique works by sampling and filtering the input signal using a set of 2-D manifolds adapted to the original data. Its cost is linear in the number of pixels and in the dimensionality of the space in which the filter operates. These techniques are significantly faster than previous approaches, supporting high-dimensional filtering of images, videos, and global illumination effects in real time. In the talk, I will present several examples illustrating their use in graphics, image, and video processing applications.Item Review and Comparative Study of Ray Traversal Algorithms on a Modern GPU Architecture(Václav Skala - UNION Agency, 2014) dos Santos, Artur Lira; Teichrieb, Veronica; Lindoso, Jorge; Skala, VáclavIn this paper we present a chronological review of five distinct data structures commonly found in literature and ray tracing systems: Bounding Volume Hierarchies (BVH), Octrees, Uniform Grids, KD-Trees, and Bounding Interval Hierarchies (BIH). This review is then followed by an extensive comparative study of six different ray traversal algorithms implemented on a modern Kepler CUDA GPU architecture, to point out pros and cons regarding performance and memory consumption of such structures. We show that a GPU KD-Tree ray traversal based on ropes achieved the best performance results. It surpasses the BVH, often used as primary structure on state-of-the-art ray tracers. Compared to BVHs, a carefully well implemented ropes-based KD-Tree is from 13 to 71% faster. However, it may consume up to 9x more memory than other structures. This disadvantage can be a limiting factor on memory limited architectures.Item A Lightweight Approach for Analyzing Insect Behavior on a Mobile System(Václav Skala - UNION Agency, 2014) Marcovecchio, Diego; Delrieux, Claudio; Werdin, Jorge; Stefanazzi, Natalia; Skala, VáclavThis paper describes the research, design, and development of a ligthweight system created in order to track multiple insects on just-recorded videos, and generate statistics using the obtained information. The system was created in order to replace the previous methods (manual monitoring that requires a lot of human assistance, or expensive specific instrumental) used in a laboratory to analyze insect behavior, and uses algorithms that are fast enough to allow it to run on a simple mobile phone. The final installation is simple and unexpensive, and allows the researchers to immediately obtain the data needed to evaluate their work. Typical segmentation and tracking problems (bad-quality frames, changes on the appearence of the tracked objects or the background scene, occlusion between an object and the scene or between multiple objects, or camera movement) were attacked. The tortuosity of the insect’s path was calculated using our own algorithm to approximate the fractal dimension of the trails, and the result is a flexible system that allows researchers to record and immediately analyze the behavior of multiple insects in a laboratory.Item Supervised Force Directed Algorithm for the Generation of Flow Maps(Václav Skala - UNION Agency, 2014) Debiasi, Alberto; Simões, Bruno; De Amicis, Raffaele; Skala, VáclavCartographic flow maps are graphical representations for portraying the movement of objects, such as people, goods or traffic network, from one location to another. On the one hand, flow maps can reduce visual clutter by merging single representations of movement. On the other hand, flow maps are also fast to produce and simple to understand. In this paper, we present a new method for the automatic generation of flow maps. Our method is based on a theoretically grounded physical system to describe the motion and forces of attraction and repulsion between data points. Additionally, support for an optional supervision of the graph layout is also available. Finally, the cost of our algorithm is evaluated and a comparison with existing implementations is provided. Results have shown a good balance between computational complexity and the visual quality of the generated maps.Item A Stable and Invariant Three-polar Surface Representation: Application to 3D Face Description(Václav Skala - UNION Agency, 2014) Jribi, Majdi; Ghorbel, Faouzi; Skala, VáclavIn this paper, we intend to introduce a new curved surface representation that we qualify by three-polar. It is constructed by the superposition of the three geodesic potentials generated from three reference points of the surface. By considering a pre-selected levels set of this superposition, invariant points are obtained. A comparative study between this representation and the unipolar one based on the level curves around one reference point is established in the sense of the stability under errors on the reference points positions. The three-polar representation is applied, finally, for 3D human faces description. Its accuracy is performed in the mean of the Hausdorff shape distance.Item Improved Estimation of Tissue Noise Power Spectra in CT Data(Václav Skala - UNION Agency, 2014) Walek, Petr; Jan, Jiří; Ouředníček, Petr; Škotáková, Jarmila; Jíra, Igor; Skala, VáclavEvaluation and measuring of image quality in X-ray computed tomographic (CT) data gained importance with recent appearance of modern algorithms for iterative reconstruction of CT data. Thanks to the ability of dramatically reducing applied radiation dose declaratively without loss of image quality, they are expected to replace the conventionally used filtered back projection (FBP) algorithm. Quality of iteratively reconstructed data in terms of image noise is routinely evaluated in images of homogeneous phantoms or in small regions of interest in real patient data. Character of the noise, whose characteristics are dependent on imaged scene, require measuring in the whole volume of real patient data and moreover in diverse tissues separately. This paper presents generalization of one dimensional noise power spectra estimation which enables its calculation from separate tissues. Firstly, basic tissues must be segmented and the resulting segmentation masks are used for the noise power spectra estimation. The estimation carried out with the help of the binary segmentation masks is, due to convolutional property of the Fourier transform, burdened by error due to spectral leakage. A binary segmentation mask may be seen as a two-dimensional windowing function with steep borders. Our method for reduction of the error is based on replacement of binary segmentation masks by designed two-dimensional spatially adaptive windowing functions with better spectral properties. Design of the spatially adaptive windows is based on distance maps and optimized skeletonization calculated using the maximal discs approach. The magnitude of the segmentation introduced error can be experimentally measured using a simulated noise with known power spectrum, which is compared with the noise power spectrum estimated in frame of the segmented tissue (i.e. affected by the spectral leakage). Finally, it is shown that the proposed two-dimensional spatially adaptive windowing functions are able to significantly improve precision of the noise power spectra estimation in diverse tissues.Item A Physically-based Simulation of a Caenorhabditis elegans(Václav Skala - UNION Agency, 2014) Mujika, Andoni; de Mauro, Alessandro; Robin, Gautier; Epelde, Gorka; Oyarzun, David; Skala, VáclavThis paper shows the three-dimensional physical model created to simulate the locomotion of the Caenorhabditis elegans. The C. elegans is a very deep studied nematode as it is considered one of the simplest nervous systems in nature, made of 302 neurons and about 8000 connections. To date, there is no system that can faithfully reproduce the rich behavioral repertoire of this tiny worm in terms of neural activity and locomotion. The Si Elegans project aims to develop the first hardware-based computing framework that will accurately mimic a C. elegans worm in real time. It will enable complex and realistic behavior to emerge through interaction with a rich and dynamic simulation of a natural or laboratory environment. As a result, the locomotion of the virtual worm will be rendered in a web-based platform. In this paper, we describe the approach followed for the physically-based modeling and simulation of C. elegans and the benefits of our approach compared to existing ones. The main contribution of our work is the utilization of biphasic springs in the structure that represents the worm in the virtual environment and a Finite Element Method based internal force field to simulate the internal pressure of the body.Item Image-based 3D Modelling in Archaeology: Application and Evaluation(Václav Skala - UNION Agency, 2014) Bernardes, Paulo; Magalhães, Fernanda; Ribeiro, Jorge; Madeira, Joaquim; Martins, Manuela; Skala, VáclavImage-based 3D modelling tools and techniques can be used to support some stages of the archaeological process. Two application examples, for two sites of the Roman Era, are presented, illustrating the usage and usefulness of such tools for the archaeological survey. For a quadrangular pool in the Milreu Roman Villa (Faro, Portugal), the 3D models resulting from the application of two different image-based modelling tools, using the same initial set of digital photos, are compared regarding time spent and model accuracy. For the Fonte do Ídolo in Bracara Augusta (Braga, Portugal), the result of a traditional survey is compared both with a laser-based survey and an image-based survey.Item Accurate thickness computation of a B-Rep model on the GPU(Václav Skala - UNION Agency, 2014) Lemasson, Grégoire; Iehl, Jean-Claude; Zara, F.; Baudet, Vincent; Arthaud, Philippe; Shariat, Behzad; Skala, VáclavIn this paper, we present a technique that calculates the thickness of a B-Rep model. This calculation is performed directly on trimmed NURBS and not on a triangular approximation. We determine the thickness in parallel on GPU, by computing the radius of maximal spheres contained within the B-Rep model. The results are presented by a color-coded thickness map. A detailed study of our results demonstrates a very important gain in stability, precision and computation time compared to others approaches.Item GPU-based Adaptive Surface Reconstruction for Real-time SPH Fluids(Václav Skala - UNION Agency, 2014) Du, Shuchen; Kanai, Takashi; Skala, VáclavWe propose a GPU-based adaptive surface reconstruction algorithm for Smoothed-Particle Hydrodynamics (SPH) fluids. The adaptive surface is reconstructed from 3-level grids as proposed by [Akinci13]. The novel part of our algorithm is a pattern based approach for crack filling, which is recognized as the most challengeable part of building adaptive surfaces. Unlike prior CPU-based approaches [Shu95, Shekhar96, Westermann99, Akinci13] that detect and fill cracks according to some criteria during program running that were slow and unrobust, all the possible crack patterns are analyzed and defined in advance and later, during program running, the cracks are detected and filled according to the patterns. Our approach is thus robust, GPU-friendly, and easy to implement. Results obtained show that our algorithm can produce surface meshes of almost the same quality as those produced by the conventional Marching Cubes method, with significantly reduced computation time and memory usage.Item A Multi-threaded Multi-context GPU Methodology for Real-time Terrain Rendering(Václav Skala - UNION Agency, 2014) Yusuf, Mohammad A.; Mostafa, Mostafa G. M.; Elarif, Taha I.; Skala, VáclavReal-time rendering of large terrains has several important applications. Hence, many methods have been devised to solve this problem. The main challenge for such methods is to deal with a large terrain dataset and maintain interactive frame rates. In this paper, we propose a level-of-detail (LOD) based multi-threaded multi-context method that works on two separate activities. Each activity is assigned to its own CPU thread and GPU context. The LOD hierarchy is constructed on the GPU context of the errors activity and stored as a 2D texture map. This texture map is used by the blocks rendering activity via its CPU thread to initiate the rendering process by sending different terrain blocks as translation and scaling parameters to its GPU context, which uses a reusable single shared vertex and index buffer to render the required block based on the passed parameters and the height-field texture. The results show that the proposed method achieves high interactive frame rates at guaranteed very small screen-space errors.Item Efficient Hand-Over Motion Reconstruction(Václav Skala - UNION Agency, 2014) Mousas, Christos; Newbury, Paul; Anagnostopoulos, Christos-Nikolaos; Skala, VáclavIn this paper, a hand-over motion reconstruction methodology is presented. First to be examined in the proposed approach is the way in which the system computes the optimal markerset for a given dataset of hand motion sequences. In a second step before the motion reconstruction process and given the reduced number of markers, the system estimates the remaining markers by computing a simple distance metric. Having the complete number of markers, including both the input and the computed markers, the system reconstructs the motion of the character’s fingers. The reconstruction process is formulated in a maximum a posteriori framework, which is responsible for approximating a valid pose of the character’s hand, in which the mixture of factor analysis (MFA) clustering techniques was used for the prior learning process. The results show that high quality motions of the character’s hand can be reconstructed with the methodology presented.Item Character Transfer: Example-based individuality retargeting for facial animations(Václav Skala - UNION Agency, 2014) Kato, Takuya; Saito, Shunsuke; Kawai, Masahide; Iwao, Tomoyori; Maejima, Akinobu; Morishima, Shigeo; Skala, VáclavA key disadvantage of blendshape animation is the labor-intensive task of sculpting blendshapes with individual expressions for each character. In this paper, we propose a novel system ”Character Transfer”, that automatically sculpts blendshapes with individual expressions by extracting them from training examples; this extraction creates a mapping that drives the sculpting process. Comparing our approach with the naïve method of transferring facial expressions from other characters, Character Transfer effectively sculpted blendshapes without the need to create such unnecessary blendshapes for other characters. Character Transfer is applicable even the training examples are limited to only a few number by using region segmentations of the face and the blending of the mappings.Item Median mixture model for background – foreground segmentation in video sequences(Václav Skala - UNION Agency, 2014) Graszka, Piotr; Skala, VáclavThe purpose of this paper is to present a novel approach to the Gaussian mixture background modeling model (GMM) that we call the median mixture model (MMM). The proposed method is based on the same principles as the GMM, but all of the background model parameters are estimated in a much more efficient way resulting in accelerating the algorithm by about 25% without deteriorating the modeling results. The second part of this paper describes a method of uniting three MMMs where three different sets of input data undergo modeling in order to achieve even better results. This approach called the united median mixtures is more robust to random noise as well as unwanted shadows and reflections. Both algorithms are thoroughly tested and compared against the Gaussian mixture model, taking into consideration robustness to noise, shadows and reflections.Item Reliable and Unobtrusive Inter-Device Collaboration by Continuous Interaction(Václav Skala - UNION Agency, 2014) Kajan, Rudolf; Szentandrási, István; Herout, Pavel; Pavelková, Alena; Skala, VáclavCurrently, there is a lack of support for seamless task migration among devices - starting a task on one device and continuing it on another, without the need of manual application state inspection and data transfer. We are solving this problem by employing our framework for application state acquisition coupled with user interface based on an intuitive metaphor: video recording. Our solution utilizes a combination of natural features based detection and marker tracking in order to reliably establish the homography between the screen and the observation of the mobile device’s camera. This allows us to employ a fast and precise continuous interaction even on low-end mobile devices. In every moment, user is given relevant task and content-migration options for selected application. The experimental results show that our solution provides reliable task migration at interactive frame rates.Item Single-Phase Trapped Air Simulation in Water Flow(Václav Skala - UNION Agency, 2014) Oh, Seungtaik; Jeong, Il Kwon; Skala, VáclavWe introduce a novel practical single-phase particle simulation for trapped air bubbles in a turbulent water flow. Our model for a trapped air bubble is a low-density rigid body with a spherical shape, and our bubble interacts with water and other rigid bodies in a fully two-way manner. Our bubble is created at a trapped air pocket computed from the water volume. Stable and realistic bubble interactions are achieved using an impulse-based boundary force with non-positive coefficients of restitution. Subgrid-scale bubbles are also created to add more details using precomputed bubble data and an oscillating bubble mesh is used in rendering stage instead of a spherical shape for a soft look of the bubble surface. Our method can be easily implemented by extending an existing rigid body interaction of fluid solver, and it is fast compared to two-phase simulation because we do not simulate the air part.Item Scalable rendering for very large meshes(Václav Skala - UNION Agency, 2014) Chajdas, Matthäus G.; Reitinger, Matthias; Westermann, Rüdiger; Skala, VáclavIn this paper, we present a novel approach for rendering of very large polygonal meshes consisting of several hundred million triangles. Our technique uses the rasterizer exclusively to allow for high-quality, anti-aliased rendering and takes advantage of a compact, voxel-based level-of-detail simplification. We show how our approach unifies streaming, occlusion culling, and level-of-detail into a single rasterization based pipeline. We also demonstrate how our level-of-detail simplification can be quickly computed, even for the most complex polygonal meshes.Item Hexagonal Image Quilting for Texture Synthesis(Václav Skala - UNION Agency, 2014) Kuri, David; Root, Elena; Theisel, Holger; Skala, VáclavThe synthesis of textures of arbitrary size from smaller samples is a much-noticed problem in the field of computer graphics. While the proposed solutions deliver very good results for regular and near-regular textures, the synthesis of irregular textures is in need of improvement. In this paper, the well-known Image Quilting algorithm is analyzed and its idea is enhanced by replacing the square shape of the patches by a hexagonal shape. In addition, rotation and mirroring of patches are introduced. A penalty map is used to enforce even usage of source data and transformations to make feature repetition less noticeable and improve synthesis from multiple source images. This leads to considerably better results for complex textures like wood, smoke or water waves.
