Computer Graphics & Visual Computing (CGVC) 2020

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2632929

Browse

Now showing 1 - 16 of 16

Breathing Life into Statues Using Augmented Reality
(The Eurographics Association, 2020) Ioannou, Eleftherios; Maddock, Steve; Ritsos, Panagiotis D. and Xu, Kai
AR art is a relatively recent phenomenon, one that brings innovation in the way that artworks can be produced and presented in real-world locations and environments. We present an AR art app, running in real time on a smartphone, that can be used to bring to life inanimate objects such as statues. The work relies on a virtual copy of the real object, which is produced using photogrammetry, as well as a skeleton rig for subsequent animation. As part of the work, we present a new diminishing reality technique, based on the use of particle systems, to make the real object 'disappear' and be replaced by the animating virtual copy, effectively animating the inanimate. The approach is demonstrated on two objects: a juice carton and a small giraffe sculpture.
CGVC 2020: Frontmatter
(The Eurographics Association, 2020) Ritsos, Panagiotis D.; Xu, Kai; Ritsos, Panagiotis D. and Xu, Kai
CLAWS: Computational Load Balancing for Accelerated Neighbor Processing on GPUs using Warp Scheduling
(The Eurographics Association, 2020) Gross, Julian; Köster, Marcel; Krüger, Antonio; Ritsos, Panagiotis D. and Xu, Kai
Nearest neighbor search algorithms on GPUs have been improving for years. Starting with tree-based approaches in the middle 70's, state-of-the-art methods use hash-based or grid-based methods. Leveraging high-performance hardware functionality decreases runtime of these search algorithms. Furthermore, memory consumption has been decreased significantly as well using Shared Memory. In the scope of these enhancements, particles have been reordered by different constraints that simplify neighbor processing. However, inspecting the existing algorithms reveals underused capabilities caused by algorithm desing. Exploiting these capabilities in a smart way can increase occupancy and efficiency on GPUs. In this paper, we present a neighbor processing approach that is based on dynamic load balancing. We rely on a lightweight workload-analysis phase that is applied during neighbor processing to distribute work throughout all warps in a thread group on-the-fly. In different domains, the neighbor function is often symmetric and, thus, commutative in each argument. In contrast to prior work, we use this domain knowledge to reduce the number of memory accesses considerably. Measurements of the newly introduced features on our evaluation scenarios show a comparable runtime performance to state-of-the-art methods. Increasing the overall workload by processing million-particle domains leads to significant improvements in terms of runtime. At the same time, we minimize global memory consumption to enable more particles to be processed compared to current approaches.
Continuous-Line-Based Halftoning with Pinwheel Tiles
(The Eurographics Association, 2020) Ahmed, Abdalla G. M.; Ritsos, Panagiotis D. and Xu, Kai
We describe a technique for continuous-line-based halftoning with space-filling curves, namely Segerman's pinwheel curve, which is based on Conway's pinwheel tiles. We adapt the length of the curve passing through the tiles in accordance with a supplied density map, e.g. a grayscale image. We use the subdivision depth of the tiling for a coarse control of the curve length, and we describe a finer control to faithfully match the density map. Our technique exhibits excellent temporal coherence, making it suitable for animated images.
Controlling Game Objects Using Multiple Degrees-Of-Freedom
(The Eurographics Association, 2020) Sandoval, Mario; Morris, Tim; Turner, Martin; Ritsos, Panagiotis D. and Xu, Kai
LISU (Library for Interactive Settings and User-modes) is an input management computing framework which enables groups of researchers to cohabit real-time simulation environments simultaneously and to visualise and manipulate virtual objects within multiple computer-assisted visualisation applications. The key novelty of LISU is an automated layered approach (physicaldriver- transport-upper layers) with importantly a built-in HCI ontology and strictly defined set of sub-APIs between the layers. All of this allows multiple input devices with multiple degrees of freedom to interact simultaneously, allowing for more intuitive and natural behaviour. Evaluation combines both linear and non-linear user modes, with a comparison system provided by Unity3D. By combining human spatial reasoning and computer graphics theory, technologies like LISU have the potential to improve our ability to understand, test and evaluate, reengineer, and then communicate better virtual dataset behaviour.
Emotion Transfer for 3D Hand Motion using StarGAN
(The Eurographics Association, 2020) Chan, Jacky C. P.; Irimia, Ana-Sabina; Ho, Edmond S. L.; Ritsos, Panagiotis D. and Xu, Kai
In this paper, we propose a new data-driven framework for 3D hand motion emotion transfer. Specifically, we first capture highquality hand motion using VR gloves. The hand motion data is then annotated with the emotion type and converted to images to facilitate the motion synthesis process and the new dataset will be available to the public. To the best of our knowledge, this is the first public dataset with annotated hand motions. We further formulate the emotion transfer for 3D hand motion as an Image-to-Image translation problem, and it is done by adapting the StarGAN framework. Our new framework is able to synthesize new motions, given target emotion type and an unseen input motion. Experimental results show that our framework can produce high quality and consistent hand motions.
Encyclopaedia-based Framework for 3D Image Processing Applications
(The Eurographics Association, 2020) Morley, Terence; Morris, Tim; Turner, Martin; Ritsos, Panagiotis D. and Xu, Kai
The uses of unmanned aerial vehicles (UAV) are rapidly increasing across diverse applications including surveillance, policing and search and rescue. To perform domain-specific functions, software systems incorporating 2D and 3D image processing libraries are being developed to work on the recorded and streamed video. But how agile are these systems? Can their operation be modified by users? How easy is it to add or replace UAVs or their preferred imaging module or improved compute resources? In this work-in-progress paper, we present an encyclopaedia-based framework (EbF) that can answer positively to these questions. Our novel EbF specifies the use of drop-in modules to enable speedy implementation and modification of systems by the operator and, as it incorporates knowledge of the input image-capture devices and presentation preferences, the system includes automated parameter selection. Central to the framework is an encyclopaedia which is used to store all information pertaining to the current system operation and can be used by imaging modules to ensure that they can adapt to changes within the system or its environment. Results are shown over three use-case implementations that are easy to control and set-up by novice operators utilising simple computational wrapper scripts.
Generalized K-means for Metric Space Clustering Using PageRank
(The Eurographics Association, 2020) Hajij, Mustafa; Said, Eyad; Todd, Robert; Ritsos, Panagiotis D. and Xu, Kai
We utilize the PageRank vector to generalize the k-means clustering algorithm to directed and undirected graphs. We demonstrate that PageRank and other centrality measures can be used in our setting to robustly compute centrality of nodes in a given graph. Furthermore, we show how our method can be generalized to metric spaces and apply it to other domains such as point clouds and triangulated meshes.
A Gesture Recognition Model for Virtual Reality Motion Controllers
(The Eurographics Association, 2020) Headleand, Chris; Williams, Benjamin; Holopainen, Jussi; Gilliam, Marlon; Ritsos, Panagiotis D. and Xu, Kai
In this paper we discuss gesture recognition in the domain of Virtual Reality (VR) video games. We begin by presenting a detailed review of the literature. Furthermore, we discuss some of the specific opportunities and challenges that are specific to the VR domain. Most commercial VR devices come with tracked motion controllers as a default interface which facilitates the possibility of gesture control. However, video games specifically require a high degree of accuracy to prevent non-gesture actions being evaluated. To tackle this challenge we present a novel modification to the Hidden Markov Model gesture recognition approach. We expand on previous work with gestures in with the implementation of an adaptive database system allowing users to quickly engage with an application without significant training. Our results on a benchmark problem shows that the approach can produce impressive accuracy rates. The results from our benchmarking shows promise for the usability of gesture based interaction systems for VR devices in the future. Our system achieves high levels of recognition accuracy competitive with the best performing existing system whilst requiring minimal user independent training.
Interaction Framework within Collaborative Virtual Environments for Multiple Users each interacting with Multiple Degrees-Of-Freedom Controllers
(The Eurographics Association, 2020) Sandoval, Mario; Morris, Tim; Turner, Martin; Ritsos, Panagiotis D. and Xu, Kai
Collaboration is a process in which two or more agents work together to achieve shared goals. However, many existing platforms cannot generate a collaborative environment to engage multiple users with multiple controllers in a seamless manner. To address this need, this poster and work in progress article will describe LISU (Library for Interactive Settings and User-modes) an input management computing framework that enables collaboration across multiple input controllers as its default. Within the system team members cohabit any real-time simulation environments simultaneously and are then able to jointly control visualisation software across multiple controllers while being continually monitored and evaluated at a low level, allowing research questions to be answered.
Medical Ultrasound Training in Virtual Reality
(The Eurographics Association, 2020) Elliman, James P.; Bethapudi, Sarath; Koulieris, George Alex; Ritsos, Panagiotis D. and Xu, Kai
In this work we propose a novel training solution for learning and practising the core psychomotor skills required in Diagnostic Ultrasound examinations with a computer-based simulator. This is in response to the long-standing challenges faced by educators in providing regular training opportunities as a shortage of equipment, staff unavailability and cost, hamper the current training model. We propose an alternative, VR-based model with a highly realistic 3D environment. To further realism of the experience, 3D printed props that work in conjunction with the simulation software will be designed. Our approach further extends previous work in generative model-based US simulation by developing a ray-tracing algorithm for use with the recently released NVidia RTX technology.
ORQA: Objective Reflection Quality Assessment
(The Eurographics Association, 2020) Houwink, Sietze G.; Kliffen, Klaas Y.; Kosinka, Jiri; Ritsos, Panagiotis D. and Xu, Kai
We present ORQA (Objective Reflection Quality Assessment), a method to objectively assess shape quality from reflection line renderings. The goal of ORQA is to correctly order existing comparable reflection line renderings according to perceived shape quality. Surface quality information is extracted from directional changes of the reflection lines. Relative importance on the directional changes and reflection line length are key aspects of scoring. ORQA is fast, stable and generalises well over various datasets.
Persistent Homology and the Discrete Laplace Operator For Mesh Similarity
(The Eurographics Association, 2020) Hajij, Mustafa; Zhang, Yunhao; Liu, Haowen; Rosen, Paul; Ritsos, Panagiotis D. and Xu, Kai
We use persistent homology along with the eigenfunctions of the Laplacian to study similarity amongst geometric and combinatorial objects. Our method relies on studying the lower-star filtration induced by the eigenfunctions of the Laplacian. This gives us a shape descriptor that inherits the rich information encoded in the eigenfunctions of the Laplacian. Moreover, the similarity between these descriptors can be easily computed using tools that are readily available in Topological Data Analysis. We provide experiments to illustrate the effectiveness of the proposed method.
Recognising Specific Foods in MRI Scans Using CNN and Visualisation
(The Eurographics Association, 2020) Gardner, Joshua; Al-Maliki, Shatha; Lutton, Évelyne; Boué, François; Vidal, Franck; Ritsos, Panagiotis D. and Xu, Kai
This work is part of an experimental project aiming at understanding the kinetics of human gastric emptying. For this purpose magnetic resonance imaging (MRI) images of the stomach of healthy volunteers have been acquired using a state-of-art scanner with an adapted protocol. The challenge is to follow the stomach content (food) in the data. Frozen garden peas and petits pois have been chosen as experimental proof-of-concept as their shapes are well defined and are not altered in the early stages of digestion. The food recognition is performed as a binary classification implemented using a deep convolutional neural network (CNN). Input hyperparameters, here image size and number of epochs, were exhaustively evaluated to identify the combination of parameters that produces the best classification. The results have been analysed using interactive visualisation. We prove in this paper that advances in computer vision and machine learning can be deployed to automatically label the content of the stomach even when the amount of training data is low and the data imbalanced. Interactive visualisation helps identify the most effective combinations of hyperparameters to maximise accuracy, precision, recall and F1 score, leaving the end-user evaluate the possible trade-off between these metrics. Food recognition in MRI scans through neural network produced an accuracy of 0.97, precision of 0.91, recall of 0.86 and F1 score of 0.89, all close to 1.
Simulating Dynamic Ecosystems with Co-Evolutionary Agents
(The Eurographics Association, 2020) Ferguson, Gary; Vidal, Franck; Ritsos, Panagiotis D. and Xu, Kai
As video games grow in complexity and require increasingly large and immersive environments, there is a need for more believable and dynamic characters not controlled by the player, known as non-player character (NPC). Video game developers will often face the challenge of designing these NPCs in a time efficient manner. We propose an agent-based Cooperative Co-evolution Algorithm (CCEA) where NPCs are implemented as artificial life (AL) agents that are created through an evolutionary process based on simple rules. The virtual environment can be filled with a range of interesting agents, each acting independently from one another, to fulfil their own wants and needs. The proposed middleware framework is suitable for computer animation of NPCs and the development of video games, especially where swarm intelligence is simulated. We proved that agents implemented with a very limited number of variables making up their genome can be successfully integrated in a co-evolutionary multi-agent system (CoEMAS). Results showed promising levels of speciation and interesting emergent and plausible behaviours amongst the agents.
Visualizing Usage Data from a Diabetes Management System
(The Eurographics Association, 2020) Duce, David A.; Martin, Clare; Russell, Alex; Brown, Dan; Aldea, Arantza; Alshaigy, Bedour; Harrison, Rachel; Waite, Marion; Leal, Yenny; Wos, Marzena; Fernandez-Balsells, Mercè; Real, José Manuel Fernández; Nita, Lucian; López, Beatriz; Massana, Joaquim; Avari, Parizad; Herrero, Pau; Jugnee, Narvada; Oliver, Nick; Reddy, Monika; Ritsos, Panagiotis D. and Xu, Kai
This article explores the role for visualization in interpreting data collected by a customised analytics framework within a healthcare technology project. It draws on the work of the EU-funded PEPPER project, which has created a personalised decision-support system for people with type 1 diabetes. Our approach was an exercise in exploratory visualization, as described by Bergeron's three category taxonomy. The charts revealed different patterns of interaction, including variability in insulin dosing schedule, and potential causes of rejected advice. These insights into user behaviour are of especial value to this field, as they may help clinicians and developers understand some of the obstacles that hinder the uptake of diabetes technology.

Browse

Browsing Computer Graphics & Visual Computing (CGVC) 2020 by Title

Results Per Page

Sort Options