Eurographics Digital Library
Not a member yet
17121 research outputs found
Sort by
NOVA-3DGS: No-reference Objective VAlidation for 3D Gaussian Splatting
No full textIn recent years, radiance field methods, and in particular 3D Gaussian Splatting (3DGS), have distinguished themselves in the field of image-based rendering and scene reconstruction techniques, gaining significant success in academia and being cited in numerous research papers. Like other methods, 3DGS requires a large and diverse dataset of images for network training as a fundamental step to ensure effectiveness and high-quality results. Consequently, the acquisition phase is highly time-consuming, especially considering that a portion of the acquired dataset is not actually used for training but is reserved for testing. This is necessary because all commonly used metrics for evaluating the quality of 3D reconstructions, such as PSNR and SSIM, are reference-based metrics; i.e., requiring a ground truth. In this work, we present NOVA, a study focused on no-reference evaluation of 3DGS renders, based on key metrics in this field: PSNR and SSIM.Eurographics 2025 - PostersPoster
Experimenting with Young Adults' Digital Engagement to Leverage Cultural Heritage as Catalyst for Sustainable Development Goals: The Open Atelier Digital Competition
No full textThe contribution aims at throwing light on the potential of an emerging digital practice - ''digital creative competitions'' - and in particular on its effectiveness in promoting innovative ways to foster audience engagement in the interpretation, dissemination, and use of cultural heritage within contemporary scenarios. These initiatives are based on the collection and awarding of original audiovisual and multimedia projects, aimed to promote educative interactions with cultural resources and formulate multiple perspectives and applications around them. This practice is observed here in the light of the experience carried out within Open Atelier, a project funded by the European Commission within the Creative Europe program. The Open Atelier Digital Competition was launched in 2024 to foster young creative talents to develop innovative reinterpretations of the collections of four European house museums, based on their possible connection with the achievement of one or multiple tasks blueprinted by the Sustainable Development Goals at the core of the United Nations' 2030 Agenda. The evaluation of the outcomes of this experience provides evidence about the strategic role of digital audience engagement in generating impactful innovations in the cultural sector. By activating online co-creative forums through which young generations become the producers and narrators of new and resonant perspectives, this format enables the challenging of institutionalized and traditional views of heritage, and contributes to rethink its role as a resource for contemporary scenarios.Digital HeritageDigital Heritage, Tourism, and Sustainabilit
A Gaze Prediction Model for Task-Oriented Virtual Reality
No full textIn this work, we present a gaze prediction model for Virtual Reality task-oriented environments. Unlike past work which focuses on gaze prediction for specific tasks, we investigate the role and potential of temporal continuity in enabling accurate predictions in diverse task categories. The model reduces input complexity while maintaining high prediction accuracy. Evaluated on the OpenNEEDS dataset, it significantly outperforms baseline methods. The model demonstrates strong potential for integration into gaze-based VR interactions and foveated rendering pipelines. Future work will focus on runtime optimization and expanding evaluation across diverse VR scenarios.Eurographics 2025 - PostersPoster
FlatCAD: Fast Curvature Regularization of Neural SDFs for CAD Models
No full textNeural signed-distance fields (SDFs) are a versatile backbone for neural geometry representation, but enforcing CAD-style developability usually requires Gaussian-curvature penalties with full Hessian evaluation and second-order differentiation, which are costly in memory and time. We introduce an off-diagonal Weingarten loss that regularizes only the mixed shape operator term that represents the gap between principal curvatures and flattens the surface. We present two variants: a finitedifference version using six SDF evaluations plus one gradient, and an auto-diff version using a single Hessian-vector product. Both converge to the exact mixed term and preserve the intended geometric properties without assembling the full Hessian. On the ABC benchmarks the losses match or exceed Hessian-based baselines while cutting GPU memory and training time by roughly a factor of two. The method is drop-in and framework-agnostic, enabling scalable curvature-aware SDF learning for engineering-grade shape reconstruction. Our code is available at https://flatcad.github.io/.Computer Graphics ForumLines, Surfaces & Fields44
A cathedral of spatialised annotations portraying the multidisciplinary study of Notre Dame de Paris
No full textThis article examines annotation in the documentation field as more than a technical feature, framing it as a structured trace of expert activity embedded in spatial, temporal, and semantic contexts. Using the large scale, multidisciplinary worksite established after the 2019-fire at Notre-Dame de Paris, as a case study, it explores how annotations function as epistemic, multiscalar, and semantically rich knowledge objects that mediate observation, interpretation, and analysis. The interdisciplinary scientific worksite provides a unique setting to test large-scale annotation practices, with hundreds of scientists from diverse disciplines converging around a shared object of study, and address challenges in tool integration, terminology, and workflows. The study focuses on semantic annotation work conducted via the aïoli platform, a web-based 3D annotation tool, analyzing a corpus of 14,000 annotations linked to over 135,000 spatialized images.Digital HeritagePredictive Analysis, AI, Simulation, and Novel Computational Method
ProTrans: Projecting In-Place Translations for Printed Text
No full textReading unknown words in non-native languages can hinder comprehension or slow reading by requiring dictionary consultation. Existing solutions for faster lookup only work digitally or, in the case of printed text, require a separate display. To enable a more seamless reading experience in the latter scenario, we present ProTrans (Projected Translation), a projector-camera system that detects words users point to and projects their translations onto nearby surfaces. We compare three projection targets-paper, finger, and hand-and gather initial user feedback. Results indicate that projecting onto the back of the hand balances legibility and viewing comfort, supporting the feasibility of skin-based projection for translation tasks.ICAT-EGVE 2025 - International Conference on Artificial Reality and Telexistence and Eurographics Symposium on Virtual EnvironmentsInterface
Interactive Visual Analytics for Local Decarbonisation Planning: Empowering Policy-Aligned Scenario Exploration
No full textDeveloping equitable and effective decarbonisation plans is a critical challenge for UK local authorities, who must balance complex technical, social, and economic factors. While computational models can propose optimal solutions based on a single objective, they often fail to account for the nuanced trade-offs and competing priorities inherent in public policy. We address this with a visual analytics system designed to support a human-in-the-loop planning process. Our primary contributions are threefold: (i) a modular, component-based planning paradigm that makes the construction of complex, multi-objective strategies cognitively manageable; (ii) a multi-scale visualisation framework that uses a model-driven glyph design to represent multivariate and temporal data uniformly across geographic scales, enabling fair and just assessment; and (iii) a tightlyintegrated workflow that allows planners to iteratively explore data, compose interventions, simulate outcomes, and refine their strategies in real-time. We demonstrate through an application scenario how our system empowers planners to move beyond monolithic optimisation and engage in a transparent, evidence-based dialogue with their data, ultimately supporting the creation of more robust and equitable decarbonisation plans.Computer Graphics and Visual Computing (CGVC)Visualisatio
"Wild West" of Evaluating Speech-Driven 3D Facial Animation Synthesis: A Benchmark Study
No full textRecent advancements in the field of audio-driven 3D facial animation have accelerated rapidly, with numerous papers being published in a short span of time. This surge in research has garnered significant attention from both academia and industry with its potential applications on digital humans. Various approaches, both deterministic and non-deterministic, have been explored based on foundational advancements in deep learning algorithms. However, there remains no consensus among researchers on standardized methods for evaluating these techniques. Additionally, rather than converging on a common set of datasets and objective metrics suited for specific methods, recent works exhibit considerable variation in experimental setups. This inconsistency complicates the research landscape, making it difficult to establish a streamlined evaluation process and rendering many cross-paper comparisons challenging. Moreover, the common practice of A/B testing in perceptual studies focus only on two common metrics and not sufficient for non-deterministic and emotion-enabled approaches. The lack of correlations between subjective and objective metrics points out that there is a need for critical analysis in this space. In this study, we address these issues by benchmarking state-of-the-art deterministic and non-deterministic models, utilizing a consistent experimental setup across a carefully curated set of objective metrics and datasets. We also conduct a perceptual user study to assess whether subjective perceptual metrics align with the objective metrics. Our findings indicate that model rankings do not necessarily generalize across datasets, and subjective metric ratings are not always consistent with their corresponding objective metrics. The supplementary video, edited code scripts for training on different datasets and documentation related to this benchmark study are made publicly available- https://galib360.github.io/face-benchmark-project/.Computer Graphics ForumFace-First for Digital Avatars44
Visualising Game Audio: Increasing Audience Engagement Through the Ambient Display of Video Game Soundscapes
No full textThis paper presents an audio system originally developed by the authors as a multiplayer, audiovisual instrument-environment, here repurposed as ViGA, a real-time visualiser for spatialised game audio. ViGA translates directional and environmental audio cues into dynamic visual forms, offering players, spectators, and audiences an intuitive and ambient graphical representation of in-game sound, and, by extension, the emergent rhythms of gameplay. Drawing on influences from video game visualisation, audiovisual synthesis, historical colour organs, visual music, music visualisers, and ambient display design, the system situates itself at the intersection of creative computing and real-time telemetry rendering. In this paper we outline the design rationale and technical architecture of the system, including its data mapping, visual language, and rendering strategies. We then present early use cases of ViGA though the video game Counter-Strike II, highlighting its potential to enhance spatial comprehension, engagement, and aesthetic experience.Computer Graphics and Visual Computing (CGVC)Short Papers Session: Games and Graphic
