Eurographics Digital Library
Not a member yet
17121 research outputs found
Sort by
Latent Interpretation for Diffusion Autoencoders via Integrated Semantic Reconstruction
No full textLatent interpretation enables controllable image editing by discovering semantic components in the latent space of generative models. While prior works have primarily focused on GANs, their limited inversion capabilities and generation quality hinder their applicability in diverse editing tasks. In this paper, we propose a new framework for latent interpretation on pretrained diffusion autoencoders, combining the editing flexibility of latent-based methods with the generation quality of diffusion models. Our key insight is to perform semantic guidance directly in the latent space, thereby avoiding costly pixel-space feedback and enabling end-to-end training. To this end, we introduce a bidirectional editing strategy and an integrated lightweight semantic autoencoder to effectively constrain semantic directions. Our method enables fine-grained and disentangled manipulation across various image editing tasks, including facial attributes, face pose, and style transfer. Extensive experiments demonstrate state-of-the-art performance in both visual quality and editing disentanglement, compared to widely-used GAN-based and diffusion-based baselines. To the best of our knowledge, this work represents a novel step toward identify explicit semantic directions in the latent space of diffusion models, complementing the research on latent interpretation beyond GANs toward more flexible and precise image editing. Our code available at https://github.com/Xenithon/LIDA.Pacific Graphics Conference Papers, Posters, and DemosImage Creation & Augmentatio
Intra-site Spatial Analysis at Sambuco Cave: Testing Digital 3D Approach for the Interpretation of Upper Palaeolithic Behavioural Strategies
No full textSambuco Cave, located in central Italy, preserves a stratified sequence from the Upper Palaeolithic to the Neolithic, making it an important archaeological site. This research combines traditional excavation data with Digital Heritage methodologies to enhance the understanding of site formation and spatial organization within the cave. A 3D GIS-based workflow integrates photogrammetric documentation, lithic analysis, and spatial modelling to reconstruct prehistoric behaviours and identify activity zones. The study highlights the importance of virtual reconstructions in clarifying stratigraphic relationships and interpreting the cave's usage. Results suggest a focused, task-specific occupation pattern, emphasizing short-term use, possibly related to hunting and toolmaking. This innovative approach provides valuable insights into Upper Palaeolithic settlement strategies and offers a new perspective on the spatial organization of prehistoric sites. At present, this paper represents an experimental 3d analytical approach presented here as a methodological test.Digital HeritageAnalysing and Documenting Sites and Landscape
A Unified Multi-scale Method for Simulating Immersed Bubbles
No full textWe introduce a novel unified mixture-based method for simulating underwater bubbles across a range of bubble scales. Our approach represents bubbles as a set of Lagrangian particles that are coupled with the surrounding Eulerian water volume. When bubble particles are sparsely distributed, each particle, typically smaller than the liquid grid voxel size, corresponds to an individual spherical bubble. As the sub-grid particles increase in local density our model smoothly aggregates them, ultimately forming connected, fully aerated volumetric regions that are properly resolved by the Eulerian grid. We complement our scheme with a continuous surface tension model, defined via the gradient of the bubbles' local volume fractions, which works seamlessly across this scale transition. Our unified representation allows us to capture a wide range of effects across different scales-from tiny dispersed sub-grid air pockets to fully Eulerian two-phase interfacial flows.Computer Graphics ForumSimulating Complex Systems: Turbulent, Crowded, and Shattered44
The Intermedit project: democratizing the access to Cultural Heritage of Antiquity through digital technologies
No full textThe InterMedit project: Mediterranean Exchange in Antiquity, a bridge between cultures, was born with the aim of creating a space of confluence to highlight the origin of the millenary connection maintained by three regions of the western Mediterranean (the Balearic Islands, the northeast of the Iberian Peninsula, and the south of France) through their archaeological heritage from Antiquity. Initially, the idea was to generate a simultaneous exhibition in the three museums through the digitization and creation of 3D printed replicas. With the outbreak of the COVID pandemic in early 2020, the project had to be reoriented and redesigned in a virtual environment to create the immersive virtual exhibition. Finally, in 2024, the project carried out the simultaneous exhibition with replicas of the most outstanding pieces and access to the immersive exhibition of the InterMedit project. This project has made it possible to innovate in the creation of museographic content as a way of integrating a greater diversity of audiences, democratizing access to heritage, enhancing its value, and raising public awareness of the importance of archaeological heritage as a link between territories and cultures.Digital HeritagePoster
IBE meets AIR: a framework for structured archaeological reasoning and digital reconstruction
No full textThe increasing use of digital tools in archaeology has vastly expanded our capacity to document and visualise the past, yet the reasoning that connects evidence to interpretation often remains implicit or undocumented. This paper addresses that gap by integrating Inference to the Best Explanation (IBE)-a philosophical model of hypothesis evaluation-into the Archaeological Interactive Report (AIR), a semantic platform for managing and visualising archaeological data. We demonstrate how this integration enables transparent, iterative, and reproducible archaeological interpretation by structuring hypotheses, linking them to evidence, and assessing them against explanatory virtues. Two case studies illustrate the potential of this framework: the systematic reconstruction of the House of the Greek Epigrams in Pompeii, and an experimental application to the Gribshunden shipwreck, where reasoning evolves alongside data collection. The result is a replicable method that formalises archaeological argumentation and embeds it directly within digital infrastructure, making interpretation not only visible, but critically assessable and reusable.Digital HeritageSemantics-driven Interaction with Digitized Heritag
Controlling Quadric Error Simplification with Line Quadrics
No full textThis work presents a method to control the output of mesh simplification algorithms based on iterative edge collapses. Traditional mesh simplification focuses on preserving the visual appearance. Despite still being an important criterion, other geometric properties also play critical roles in different applications, such as triangle quality for computations. This motivates our work to stay under the umbrella of the popular quadric error mesh simplification, while proposing different ways to control the simplified mesh to possess other geometric properties. The key ingredient of our work is another quadric error, called line quadrics, which can be seamlessly added to the vanilla quadric error metric. We show that, theoretically and empirically, adding our line quadrics can improve the numerics and encourage the simplified mesh to have uniformly distributed vertices. If we spread the line quadric adaptively to different regions, it can easily lead to soft preservation of feature vertices and edges. Our method is simple to implement, requiring only a few lines of code change on top of the original quadric error simplification, and can lead to a variety of user controls.Computer Graphics ForumMesh Processing44
Publishing and reusing Collections as Data across GLAM Labs and infrastructures: from principles to practice
No full textGLAM (Galleries, Libraries, Archives and Museums) institutions have been exploring the publication and reuse of digital collections suitable for computational use, inspired by recent initiatives, such as Collections as Data, which facilitate the development of best practices and guidelines. This workshop will introduce relevant aspects concerning the publication of machine-actionable collections as well as include a hands-on session to create awareness of computational access in the context of international movements such as the International GLAM Labs Community. It will also include an overview of new platforms and environments to share and reuse data.Digital HeritageWorkshop
Data Collection and Reuse in Digital Heritage: Approaches and Results from the Digital Cultural Heritage Cluster
No full textThis paper presents the Digital Cultural Heritage Cluster, uniting six Horizon Europe-funded projects to develop interoperable digital workflows for the sustainable preservation and valorization of cultural assets.Digital HeritageWorkshop
Hacking Light in the Digitization of Archaeological Glass Vessels: the Quest for Geometric Rigor
No full textThe 3D digitization of heritage has revolutionized archaeological research and the dissemination of archaeological materiality. Emerging new technologies not only allow to analyze data from new perspectives and dimensions, in a literal sense; but have also democratized access to material culture and enriched the public's experience of history. However, a significant lacuna exists in virtual collections and databases: glass objects. In both virtual museum expositions and archaeological research, ancient glass tableware is often largely forgotten. In museums, glass artifacts are often confined to display cases. In archaeological research, ancient glass is a subject that has received comparatively little scholarly attention, and the literature on its craftmanship is scant when compared to that of its functional counterparts in other materials. Fragile, fragmentary and poorly studied, ancient glassware current status can be attributed, in part, to the technical challenges posed by the 3D digitization of glass artifacts. In this way, the new computational tools for analysing archaeological data cannot access these types of productions because they are not available in a digital version. The inherent nature of glass presents significant difficulties for digitization techniques, as the optical properties of reflectance, reflexivity, and transparency prevent its surfaces from being read by any light-based method. Glass exhibits complex and diverse light-redirecting and -absorbing characteristics. Consequently, various 3D scanning and photogrammetry techniques encounter problems in geometrically capturing glass objects. The present study aims to overcome the difficulties of 3D glass acquisition by proposing the use of a new methodology based on photogrammetry, which enables the digitization of the geometry of these elusive pieces with rigor. This technique has been specifically designed to capture light at its optimal deflection in glass vessel surfaces and to record it as reliable image data in order to build geometrically accurate 3D models. Thus, this paper intends to tackle the 3D acquisition of the archaeological glass objects even beyond the renders; that is: data has been collected and processed in a hermetic and thorough polygonal mesh that can be texturized and optimized afterward. The methodology has been implemented in the analysis of Roman glassware. The resulting precise meshes make it a technique that can be applied to any computational morphometric analysis on any type of archaeological glass artifacts, in addition to completing the digital heritage collections.Digital HeritageDigitization and Technolog
Joint Deblurring and 3D Reconstruction for Macrophotography
No full textMacro lens has the advantages of high resolution and large magnification, and 3D modeling of small and detailed objects can provide richer information. However, defocus blur in macrophotography is a long-standing problem that heavily hinders the clear imaging of the captured objects and high-quality 3D reconstruction of them. Traditional image deblurring methods require a large number of images and annotations, and there is currently no multi-view 3D reconstruction method for macrophotography. In this work, we propose a joint deblurring and 3D reconstruction method for macrophotography. Starting from multi-view blurry images captured, we jointly optimize the clear 3D model of the object and the defocus blur kernel of each pixel. The entire framework adopts a differentiable rendering method to self-supervise the optimization of the 3D model and the defocus blur kernel. Extensive experiments show that from a small number of multi-view images, our proposed method can not only achieve high-quality image deblurring but also recover high-fidelity 3D appearance.Computer Graphics ForumReconstruction from Close-up Image44