Eurographics Digital Library
Not a member yet
17121 research outputs found
Sort by
Toward General-Purpose Monte Carlo PDE Solvers for Graphics Applications
No full textThis thesis develops novel Monte Carlo methods for solving a wide range of partial differential equations (PDEs) relevant to computer graphics. While traditional discretization-based approaches efficiently compute global solutions, they often require expensive global solves even when only local evaluations are needed, and can struggle with complex or fine-scale geometries. Monte Carlo methods based on the classical Walk on Spheres (WoS) approach [Muller 1956] offer pointwise evaluation with strong geometric robustness, but in practice, their application has been largely limited to interior Dirichlet problems in volumetric domains. We significantly broaden this scope by designing versatile Monte Carlo solvers that handle a diverse set of PDEs and boundary conditions, validated through comprehensive experimental results.
First, we introduce the Walk on Boundary (WoB) method [Sabelfeld 1982, 1991] to graphics. While retaining WoS’s advantages, WoB applies to a broader range of second-order linear elliptic and parabolic PDE problems: various boundary conditions (Dirichlet, Neumann, Robin, and mixed) in both interior and exterior domains. Because WoB is based on boundary integral formulations, its structure more closely parallels Monte Carlo rendering than WoS, enabling the application of advanced variance reduction techniques. We present WoB formulations for elliptic Laplace and Poisson equations, time-dependent diffusion problems, and develop a WoB solver for vector-valued Stokes equations. Throughout, we discuss how sampling and variance reduction methods from rendering can be adapted to WoB.
Next, we address the nonlinear Navier-Stokes equations for fluid simulation, whose complexity challenges Monte Carlo techniques. Employing operator splitting, we separate nonlinear terms and solve the remaining linear terms with pointwise Monte Carlo solvers. Recursively applying these solvers with timestepping yields a spatial-discretization-free method. To deal with the resulting exponential computational cost, we also propose cache-based alternatives. Both vorticity- and velocity-based formulations are explored, retaining the advantages of Monte Carlo methods, including geometric robustness and variance reduction, while integrating traditional fluid simulation techniques.
We then propose Projected Walk on Spheres (PWoS), a novel solver for surface PDEs, inspired by the Closest Point Method. PWoS modifies WoS by projecting random walks onto the surface manifold at each step, preserving geometric flexibility and discretization-free, pointwise evaluation. We also adapt a noise filtering technique for WoS to improve PWoS.
Finally, we outline promising future research directions for Monte Carlo PDE solvers in graphics, including concrete proposals to enhance WoB.EG Graphics Dissertation Onlin
Modern Digitization and 3D Visualization Technologies for Virtual Access to Hard-to-reach or Perished Historical Sites
No full textNowadays, an important issue is both the digitization of Cultural Heritage objects and the techniques of their virtual dissemination, regardless of the current state of the digitized object. This is crucial when the digitized location is at risk of closing, degradation, or complete loss due to natural disasters, armed conflicts, or political and economic conditions. Obtaining a digital twin of such an object allows not only to maintain the most accurate representation possible but also to implement an unlimited number of sessions to disseminate such an object, regardless of whether the object is only difficult to access or has already been excluded from display, damaged, or even completely lost. The paper presents detailed case studies of 3D scanning and model sharing for inner chambers of the Golden Mosque in Samarkand (Uzbekistan), the Church of the Archangel Michael in Chicago (USA), and the Church of St. Adalbert in Chicago (USA). The acquired data is based on TLS scans and photographs of the interior. The point clouds and mesh models generated from them provide a density adequate to the detail of the interior appearance and geometry. The Golden Mosque is a part of the Tillia-Kori madrasa (dating from the 17th century) located in the Registan complex in Samarkand, Uzbekistan - located in Central Asia. In turn, the Church of the Archangel Michael in Chicago (built in 1909 by the Polish diaspora) was desacralized in 2020 and closed. Its digitally documented condition shows the abandonment of the building that served the congregation until the very last moment. The demolition process of the Church of St. Adalbert in Chicago began in 2023. At the beginning of this process, it was possible to carry out a 3D scan of its interior and separately of the Pietà sculpture (a copy of Michelangelo's Madonna della Pietà) already transferred to another church. All of the above results are publicly available in the form of basic and online 3D models, panoramas, videos, and VR files.Digital HeritageDigitization and 3D Visualization for Heritage Accessibilit
3D Digitisation for Geological and Paleontological Specimens: Challenges and Solutions
No full textThis study outlines the methodology employed for the digitisation and visualisation in different 3D Virtual Environment (both web-based and desktop-based) of paleontological and geological specimens from the Geological Collection ''Giovanni Capellini Museum'', part of the University of Bologna's museum system.The digitisation process involved 87 objects of varying sizes, shapes, materials, and conservation states, following the proven methodology implemented within the Digital Twin of L'Altro Rinascimento's temporary exhibition, developed within the PNRR project CHANGES (Spoke 4). This paper describes the pipeline adopted to address specific challenges encountered during the acquisition campaign of the diverse specimen. A preliminary assessment was conducted to determine the most suitable technique for each object based on its size, shape, material, and conservation state. The study showcases different integrated techniques, including photogrammetry, structured light scanning (SLS), and terrestrial laser scanning (TLS). The results demonstrate the flexibility and reproducibility of this approach, making it suitable for different heritage preservation and dissemination applications.Digital HeritageDigital Technologies for CHANGES (CHANGES SESSION) - Part
MatplotAlt: A Python Library for Adding Alt Text to Matplotlib Figures in Computational Notebooks
No full textWe present MatplotAlt, an open-source Python package for easily adding alternative text to Matplotlib fgures. MatplotAlt equips Jupyter notebook authors to automatically generate and surface chart descriptions with a single line of code or command, and supports a range of options that allow users to customize the generation and display of captions based on their preferences and accessibility needs. Our evaluation indicates that MatplotAlt's heuristic and LLM-based methods to generate alt text can create accurate long-form descriptions of both simple univariate and complex Matplotlib fgures. We fnd that state-of-the-art LLMs still struggle with factual errors when describing charts, and improve the accuracy of our descriptions by prompting GPT4-turbo with heuristic-based alt text or data tables parsed from the Matplotlib fgure.Computer Graphics ForumInclusive Visualizatio
BoxFusion: Reconstruction-Free Open-Vocabulary 3D Object Detection via Real-Time Multi-View Box Fusion
No full textOpen-vocabulary 3D object detection has gained significant interest due to its critical applications in autonomous driving and embodied AI. Existing detection methods, whether offline or online, typically rely on dense point cloud reconstruction, which imposes substantial computational overhead and memory constraints, hindering real-time deployment in downstream tasks. To address this, we propose a novel reconstruction-free online framework tailored for memory-efficient and real-time 3D detection. Specifically, given streaming posed RGB-D video input, we leverage Cubify Anything as a pre-trained visual foundation model (VFM) for single-view 3D object detection, coupled with CLIP to capture open-vocabulary semantics of detected objects. To fuse all detected bounding boxes across different views into a unified one, we employ an association module for correspondences of multi-views and an optimization module to fuse the 3D bounding boxes of the same instance. The association module utilizes 3D Non-Maximum Suppression (NMS) and a box correspondence matching module. The optimization module uses an IoU-guided efficient random optimization technique based on particle filtering to enforce multi-view consistency of the 3D bounding boxes while minimizing computational complexity. Extensive experiments on CA-1M and ScanNetV2 datasets demonstrate that our method achieves state-of-the-art performance among online methods. Benefiting from this novel reconstruction-free paradigm for 3D object detection, our method exhibits great generalization abilities in various scenarios, enabling real-time perception even in environments exceeding 1000 square meters.Computer Graphics ForumDetecting & Estimating from images and videos44
Automatic Image-Based Coral Polyp Analysis through Multi-View Instance Segmentation
No full textWe present an automated framework for counting and measuring the polyps of Cladocora caespitosa, a Mediterranean reefbuilding coral. To our knowledge, the most practical method for counting polyps currently involves ecologists' visual inspection of a 3D model. However, measuring polyps from the model can lead to inaccuracies due to distortions in the reconstruction. Our method integrates deep learning-based instance segmentation on 2D images with 3D models for unique polyp identification, ensuring precise biometric extraction. The proposed pipeline automates polyp detection, counting, and measurement while overcoming the limitations of manual in situ methods. Laboratory validation demonstrates its accuracy and efficiency, paving the way for scalable, high-resolution phenotyping, and field monitoring of Mediterranean coral populations.Eurographics 2025 - PostersPoster
4-LEGS: 4D Language Embedded Gaussian Splatting
No full textThe emergence of neural representations has revolutionized our means for digitally viewing a wide range of 3D scenes, enabling the synthesis of photorealistic images rendered from novel views. Recently, several techniques have been proposed for connecting these low-level representations with the high-level semantics understanding embodied within the scene. These methods elevate the rich semantic understanding from 2D imagery to 3D representations, distilling high-dimensional spatial features onto 3D space. In our work, we are interested in connecting language with a dynamic modeling of the world. We show how to lift spatio-temporal features to a 4D representation based on 3D Gaussian Splatting. This enables an interactive interface where the user can spatiotemporally localize events in the video from text prompts. We demonstrate our system on public 3D video datasets of people and animals performing various actions.Computer Graphics ForumSplat-tacular Radiance Fields44
Towards Met.3D version 2: Creating a community research software for interactive 3-D visualization of meteorological data
No full textVisualization is an important and ubiquitous tool in the daily work of weather forecasters and atmospheric researchers to analyse data from simulations and observations. The domain-specific meteorological visualization tool Met.3D is an opensource effort to make interactive, 3D, feature-based, and ensemble visualization techniques accessible to the meteorological community, and at the same time to provide a framework for visualization research with application to meteorology. Since the public release of version 1.0 in 2015, Met.3D has been used in multiple visualization research projects, and has evolved into a feature-rich visual analysis tool facilitating rapid exploration of gridded atmospheric data. In this demo, we will present the current state of our efforts to contribute a new version 2 of Met.3D to the community, with extended funcionality and improved usability for meteorological users, and with a well documented implementation for extension by visualization researchers.EuroVis 2025 - PostersPosters and Demo
Evaluating the Effect of Multimodal Scenario Cues in an LLM-Supported Auditory VR Design System for Exposure Therapy
No full textPost-Traumatic Stress Disorder (PTSD) is a prevalent disorder triggered by life-threatening trauma, and exposure therapy, which involves confronting traumatic stimuli, has been proven to be highly effective for treating PTSD. Some methods have been proposed that present patients' traumatic situations using spatial sound (Auditory VR) to provide a sense of realism during therapy (Auditory VR Exposure Therapy). However, the Auditory VR used in therapy need to be tailored for each patient, and conventionally they have been manually produced by third parties, resulting in long delays before Auditory VR exposure therapy can begin. In the previous work, they developed a system that enables the creation of auditory stimuli through text-only interaction, allowing clinicians and patients to generate sounds without the involvement of third parties. However, creating Auditory VR solely through natural language interaction proved challenging, leading to issues in usability and sound quality. In this study, while maintaining the basic approach of text-based Auditory VR generation, we developed a system that enables multimodal interaction by combining text with auditory presentation of scenarios and visual presentation of spatial designs. An evaluation experiment with participants with backgrounds in medicine or healthcare demonstrated that, compared to the previous system, the system improved usability and, based on subjective evaluations, achieved significantly higher overall sound quality ratings.ICAT-EGVE 2025 - International Conference on Artificial Reality and Telexistence and Eurographics Symposium on Virtual EnvironmentsLarge Language Model (LLM