Eurographics Digital Library
Not a member yet
17121 research outputs found
Sort by
Cardioid Caustics Generation with Conditional Diffusion Models
No full textDespite the latest advances in generative neural techniques for producing photorealistic images, they lack generation of multi-bounce, high-frequency lighting effect like caustics. In this work, we tackle the problem of generating cardioid-shaped reflective caustics using diffusion-based generative models. We approach this problem as conditional image generation using a diffusion-based model conditioned with multiple images of geometric, material and illumination information as well as light property. We introduce a framework to fine-tune a pre-trained diffusion model and present results with visually plausible caustics.Eurographics 2025 - Short PapersShort Paper
Anisotropic Gauss Reconstruction and Global Orientation with Octree-based Acceleration
No full textUnoriented surface reconstruction is an important task in computer graphics. Recently, methods based on the Gauss formula or winding number have achieved state-of-the-art performance in both orientation and surface reconstruction. The Gauss formula or winding number, derived from the fundamental solution of the Laplace equation, initially found applications in calculating potentials in electromagnetism. Inspired by the practical necessity of calculating potentials in diverse electromagnetic media, we consider the anisotropic Laplace equation to derive the anisotropic Gauss formula and apply it to surface reconstruction, called ''anisotropic Gauss reconstruction''. By leveraging the flexibility of anisotropic coefficients, additional constraints can be introduced to the indicator function. This results in a stable linear system, eliminating the need for any artificial regularization. In addition, the oriented normals can be refined by computing the gradient of the indicator function, ultimately producing high-quality normals and surfaces. Regarding the space/time complexity, we propose an octree-based acceleration algorithm to achieve a space complexity of O(N) and a time complexity of O(NlogN). Our method can reconstruct ultra-large-scale models (exceeding 5 million points) within 4 minutes on an NVIDIA RTX 4090 GPU. Extensive experiments demonstrate that our method achieves state-of-the-art performance in both orientation and reconstruction, particularly for models with thin structures, small holes, or high genus. Both CuPy-based and CUDA-accelerated implementations are made publicly available at https://github.com/mayueji/AGR.Computer Graphics ForumReconstruction44
Pixels2Points: Fusing 2D and 3D Features for Facial Skin Segmentation
No full textFace registration deforms a template mesh to closely fit a 3D face scan, the quality of which commonly degrades in non-skin regions (e.g., hair, beard, accessories), because the optimized template-to-scan distance pulls the template mesh towards the noisy scan surface. Improving registration quality requires a clean separation of skin and non-skin regions on the scan mesh. Existing image-based (2D) or scan-based (3D) segmentation methods however perform poorly. Image-based segmentation outputs multi-view inconsistent masks, and they cannot account for scan inaccuracies or scan-image misalignment, while scan-based methods suffer from lower spatial resolution compared to images. In this work, we introduce a novel method that accurately separates skin from non-skin geometry on 3D human head scans. For this, our method extracts features from multi-view images using a frozen image foundation model and aggregates these features in 3D. These lifted 2D features are then fused with 3D geometric features extracted from the scan mesh, to then predict a segmentation mask directly on the scan mesh. We show that our segmentations improve the registration accuracy over pure 2D or 3D segmentation methods by 8.89% and 14.3%, respectively. Although trained only on synthetic data, our model generalizes well to real data.Eurographics 2025 - Short PapersShort Paper
Implicit UVs: Real-time Semi-global Parameterization of Implicit Surfaces
No full textImplicit representations of shapes are broadly used in computer graphics since they offer many valuable properties in design, modeling, and animation. However, their implicit and volumetric nature makes applying 2D textures fundamentally challenging. We propose a method to compute point-wise and parallelizable semi-global parameterizations of implicit surfaces for texturing, rendering, and modeling purposes. Our method not only defines local patches of parameterization, but also enables the merging of multiple adjacent patches into large and spatially coherent ones that conform to the geometry. Implemented in shaders into a sphere-tracing pipeline, our method allows users to edit the uv-fields with real-time visualization. We demonstrate how to add rendering details (texture, normal, displacement, etc.) using our parameterization, as well extending modeling tools with implicit shell maps. Furthermore, the textured objects remain implicit and can still be used in a modeling pipeline.Computer Graphics ForumThe Shape of Rendering44
Artist-Inator: Text-based, Gloss-aware Non-photorealistic Stylization
No full textLarge diffusion models have made a remarkable leap synthesizing high-quality artistic images from text descriptions. However, these powerful pre-trained models still lack control to guide key material appearance properties, such as gloss. In this work, we present a threefold contribution: (1) we analyze how gloss is perceived across different artistic styles (i.e., oil painting, watercolor, ink pen, charcoal, and soft crayon); (2) we leverage our findings to create a dataset with 1,336,272 stylized images of many different geometries in all five styles, including automatically-computed text descriptions of their appearance (e.g., ''A glossy bunny hand painted with an orange soft crayon''); and (3) we train ControlNet to condition Stable Diffusion XL synthesizing novel painterly depictions of new objects, using simple inputs such as edge maps, hand-drawn sketches, or clip arts. Compared to previous approaches, our framework yields more accurate results despite the simplified input, as we show both quantitative and qualitatively.Computer Graphics ForumStylization and Image Processing44
Automated Refined Comic Generation: From Investigation Provenance to Data Comics using Visual Narrative Structure
No full textVisual analytics has become an important approach for criminal investigations due to the increasing amount of physical and digital data related to cases. Although state-of-the-art tools are used daily to search for evidence in the data and report the investigator's findings, building such reports remains a labor-intensive manual process. Furthermore, these reports commonly contain only a manually selected set of the investigation results, but not how these results were derived. This lack of information about the chain of evidence not only weakens reproducibility and transparency, but also makes the evidence vulnerable by jurists in court. Instead of textual reports we believe annotated visuals of the actual data exploration process better portray what the investigators did and how they came to the evidence. To this end, we introduce ARC, a framework for automatically generating comic summaries for digital investigations based on the Visual Narrative Structure from comic theory. Especially, ARC is the first framework that fully automatically generates and refines comic summaries based on interactions with investigation tools.EuroVis 2025 - PostersPoster
2D and 3D Semantic Segmentation for Interpreting and Understanding 3D Heritage Spaces
No full textThe 3D digitization of Cultural Heritage (CH) sites has become increasingly requested for documentation, preservation, and analysis applications. Beyond capturing 3D spatial geometry, the semantic interpretation and understanding of digital models are critical for enabling meaningful CH studies and facilitating informed conservation strategies. However, manual annotation and classification of architectural elements and surface pathologies remain labor-intensive and time-consuming, underscoring the need for automated approaches. This study presents a comparative analysis between two distinct semantic segmentation frameworks: (1) a 2D-to-3D pipeline that projects 2D image-based detections onto 3D point clouds produced with V-SLAM data and (2) direct segmentation methods of 3D point clouds acquired with portable LiDAR sensors. These frameworks are evaluated on data acquired using two distinct mobile mapping systems (MMS): (1) a fisheye multi-camera Visual SLAM-based portable system (ATOM-ANT3D) for the 2D-to-3D pipeline; (2) a LiDAR-based MMS (Heron MS Twin Color) for the 3D segmentation methods. Achieved results demonstrate the ability of the proposed frameworks to generate semantically enriched 3D heritage data, with the 2D-to-3D method slightly outperforming the 3D segmentation techniques.Digital HeritageDigitization and Segmentatio
Implementing Curiosity Hooks and Caring Practices in the Reconstruction of Lost Polychromy: Design Prototypes for Interactive Experiences.
No full textEngaging museum audiences with problems such as the conservation of monuments or the fading or lost colours of our Cultural Heritage remains a challenge, as traditional approaches often fail to establish lasting connections. Recent museological debates highlight a ''sense of care'' as a key perspective in fostering relationships between citizens and Cultural Heritage. The PERCEIVE Horizon project explores ''caring prototypes'' to enhance engagement with coloured collections. Building on research on curiosity-driven engagement, this study investigates design strategies for digital and hybrid prototypes, specifically tailored to engage visitors with the lost polychromy of ancient statuary. Our qualitative user research in museums and educational contexts reveals that audience feel to be involved most when actively solicited by an investigative process, rather than being treated as passive observer of reconstructions. We propose here a User eXperience (UX), interface (UI) and interaction (IxD) solutions inspired by a ''care'' theory. The theory is based on three concepts: ''care practice'', ''care as a process'' and ''effort''. The result is a modular interactive UX that guides visitors through the step-by-step reconstruction of lost polychromy, integrating archaeological, literary, and scientific data to foster a deeper connection with these fragile collections.Digital HeritagePERCEIVE: Exhibiting the ''Unexhibitable'
Learning to Predict Aboveground Biomass from RGB Images with 3D Synthetic Scenes
No full textForests play a critical role in global ecosystems by supporting biodiversity and mitigating climate change via carbon sequestration. Accurate aboveground biomass (AGB) estimation is essential for assessing carbon storage and wildfire fuel loads, yet traditional methods rely on labor-intensive field measurements or remote sensing approaches with significant limitations in dense vegetation. In this work, we propose a novel learning-based method for estimating AGB from a single ground-based RGB image. We frame this as a dense prediction task, introducing AGB density maps, where each pixel represents tree biomass normalized by the plot area and each tree's image area. We leverage the recently introduced synthetic 3D SPREAD dataset, which provides realistic forest scenes with per-image tree attributes (height, trunk and canopy diameter) and instance segmentation masks. Using these assets, we compute AGB via allometric equations and train a model to predict AGB density maps, integrating them to recover the AGB estimate for the captured scene. Our approach achieves a median AGB estimation error of 1:22kg=m2 on held-out SPREAD data and 1:94kg=m2 on a real-image dataset. To our knowledge, this is the first method to estimate aboveground biomass directly from a single RGB image, opening up the possibility for a scalable, interpretable, and cost-effective solution for forest monitoring, while also enabling broader participation through citizen science initiatives.Smart Tools and Applications in Graphics - Eurographics Italian Chapter ConferenceLearning-based Algorithm
Safeguarding the Past: Blending Digital and Traditional Tools to Combat Illicit Excavations and Cultural Property Trafficking
No full textIllicit excavations and the trafficking of cultural property remain serious transnational threats, often linked to organized crime and conflict-related instability. Addressing this challenge requires interdisciplinary collaboration and the integration of innovative digital tools. The Safeguarding the Past workshop, held within the REVITALISER project activities, explores the use of Earth Observation, AI, and 3D technologies alongside legal and criminological frameworks. It brings together researchers, professionals, and authorities to strengthen cultural heritage protection.Digital HeritageWorkshop