Eurographics Digital Library
Not a member yet
17121 research outputs found
Sort by
Infrared Reflectography of the Madonna con Bambino, San Pietro Martire e San Giovannino by Lorenzo Lotto
No full textThis study presents the results of the diagnostic imaging analyses conducted on the panel painting Madonna con Bambino, San Pietro Martire e San Giovannino by Lorenzo Lotto, specifically infrared (IR) reflectography and false-color infrared (IRFC) imaging. These investigations allowed for the examination of the layers beneath the pictorial surface and provided preliminary insights into the pigments employed by the artist. Infrared reflectography revealed previously unseen images of the underdrawing, uncovering significant compositional changes. False-colour infrared imaging contributed to the reconstruction of the artist's palette, confirming the use of high-quality pigments, indicative of a prestigious commission. The integration of non-invasive diagnostic techniques and historical-artistic research provided essential insights into the artist's painting method, contributing - through a multidisciplinary approach - to the interpretation of the first dated artwork within Lorenzo Lotto's pictorial corpus.Digital HeritageInstrumental and Computational Approaches for CH Conservation and Restauratio
GenAI-Based Reconstruction of Prehistoric Remains
No full textGenerative artificial intelligence (GenAI) has emerged as a powerful tool in various fields, including archaeology. However, its application in reconstructing prehistoric archaeology presents unique challenges and limitations that warrant careful consid-eration. Realizing the potential of Generative Artificial Intelligence to archaeological reconstruction requires a nuanced ap-proach that acknowledges and addresses the inherent limitations of AI in this context. The "black box" nature of some AI algorithms can make it difficult to understand how reconstructions are generated. This lack of transparency poses challenges for scientific reproducibility and peer review in archaeological research. By adopting a theoretical framework that combines technological innovation with rigorous archaeological methodology and ethical considerations, we can work towards more accurate, inclusive, and responsible reconstructions of prehistoric societies. This framework not only addresses the current limitations of AI in prehistoric archaeology but also sets the stage for future research directions. As AI technology continues to evolve, ongoing critical evaluation and adaptation of these approaches will be essential to ensure that generative AI becomes a valuable tool in expanding our knowledge of prehistoric human societies while respecting the complexity and cultural sensi-tivity of archaeological interpretation.Digital HeritageAI and Generative Techniques for Heritage Reconstructio
ArTLLaMA: Adaptating LLaMA to Performative Art Applications
No full textPerformative Arts represent a compelling and underexplored domain for the application of Generative AI, given their rich conceptual complexity and cultural depth. This paper presents ArTLLaMA, a domain-adapted version of the LLaMA language model, designed to support natural language querying of ArTBase, the first national database of Italian theatres and theatre archives. We focus on the Text-to-SQL task: automatically translating user questions into executable SQL queries. Off-the-shelf models often fail in this setting due to a lack of domain knowledge and schema awareness. To bridge this gap, we propose a two-stage fine-tuning methodology: first, we train the model to internalize the Entity-Relationship (ER) schema of ArTBase; then, we fine-tune it on a curated set of over 800 natural language-SQL query pairs reflecting real use cases in the domain. Our results show that schema-informed fine-tuning significantly boosts accuracy, with the best model achieving over 70% exact match andgenerating correct SQL even for complex queries involving multi-table joins and aggregations. Compared to general purpose models like ChatGPT, our approach yields more accurate, schema-compliant outputs. Beyond technical improvements, this work underscores the value of interdisciplinary collaboration: by embedding domain knowledge from the humanities into AI systems, we enable new forms of access, interaction, and understanding of cultural heritage data.Digital HeritageDigitization and Segmentatio
CoSketcher: Collaborative and Iterative Sketch Generation with LLMs under Linguistic and Spatial Control
No full textSketching serves as both a medium for visualizing ideas and a process for creative iteration. While early neural sketch generation methods rely on category-specific data and lack generalization and iteration capability, recent advances in Large Language Models (LLMs) have opened new possibilities for more flexible and semantically guided sketching. In this work, we present CoSketcher, a controllable and iterative sketch generation system that leverages the prior knowledge and textual reasoning abilities of LLMs to align with the creative iteration process of human sketching. CoSketcher introduces a novel XML-style sketch language that represents stroke-level information in structured format, enabling the LLM to plan and generate complex sketches under both linguistic and spatial control. The system supports visual appealing sketch construction, including skeleton-contour decomposition for volumetric shapes and layout-aware reasoning for object relationships. Through extensive evaluation, we demonstrate that our method generates expressive sketches across both in-distribution and out-of-distribution categories, while also supporting scene-level composition and controllable iteration. Our method establishes a new paradigm for controllable sketch generation using off-the-shelf LLMs, with broad implications for creative human-AI collaboration.Pacific Graphics Conference Papers, Posters, and DemosInteraction & Virtual Realit
Advanced digitisation and AI-powered data processing for Cultural Heritage: the HERITALISE Project
No full textThe digitisation of cultural heritage assets ensures an accurate digital archive for future generations and serves as a powerful tool for conveying the knowledge and significance of material heritage to the broader public. This contribution presents the overall goals of the HERITALISE project and the foreseen activities, which will combine AI tools for data processing and metadata and paradata creation. NeRF, 3D Gaussian Splatting and LLMs will be involved in the project with different aims ensuring the advancement of digitisation methodologies and standardisation in the cultural heritage field.Digital HeritagePoster
The GDH LidArc Initiative - Pioneering Global Under-Canopy Archaeology by LiDAR
No full textThe GDH LidArc Initiative aims to support underfunded projects in Europe, Africa, Asia, and the Americas (where possible) in the acquisition, processing, and interpretation of high-quality LiDAR data for archaeology. The research team is a consortium of universities and institutes based in Italy and the USA, with leading expertise in landscape and digital archaeology, including high-resolution LiDAR. The broader goal of the initiative is to select projects that reduce inequalities in access and lead to significant discoveries. Interested groups worldwide are invited to submit proposals for research projects, which GDH and the consortium will review. The consortium will then carry out the flights, analysis, and provide the data to the project.Digital HeritageWorkshop
2D Neural Fields with Learned Discontinuities
No full textEffective representation of 2D images is fundamental in digital image processing, where traditional methods like raster and vector graphics struggle with sharpness and textural complexity, respectively. Current neural fields offer high fidelity and resolution independence but require predefined meshes with known discontinuities, restricting their utility. We observe that by treating all mesh edges as potential discontinuities, we can represent the discontinuity magnitudes as continuous variables and optimize. We further introduce a novel discontinuous neural field model that jointly approximates the target image and recovers discontinuities. Through systematic evaluations, our neural field outperforms other methods that fit unknown discontinuities with discontinuous representations, exceeding Field of Junction and Boundary Attention by over 11dB in both denoising and super-resolution tasks and achieving 3.5× smaller Chamfer distances than Mumford-Shah-based methods. It also surpasses InstantNGP with improvements of more than 5dB (denoising) and 10dB (super-resolution). Additionally, our approach shows remarkable capability in approximating complex artistic and natural images and cleaning up diffusion-generated depth maps.Computer Graphics ForumDrawn to Detail: Sketch-Based Modeling and Non-Photorealistic Rendering44
D-NPC: Dynamic Neural Point Clouds for Non-Rigid View Synthesis from Monocular Video
No full textDynamic reconstruction and spatiotemporal novel-view synthesis of non-rigidly deforming scenes recently gained increased attention. While existing work achieves impressive quality and performance on multi-view or teleporting camera setups, most methods fail to efficiently and faithfully recover motion and appearance from casual monocular captures. This paper contributes to the field by introducing a new method for dynamic novel view synthesis from monocular video, such as casual smartphone captures. Our approach represents the scene as a dynamic neural point cloud, an implicit time-conditioned point distribution that encodes local geometry and appearance in separate hash-encoded neural feature grids for static and dynamic regions. By sampling a discrete point cloud from our model, we can efficiently render high-quality novel views using a fast differentiable rasterizer and neural rendering network. Similar to recent work, we leverage advances in neural scene analysis by incorporating data-driven priors like monocular depth estimation and object segmentation to resolve motion and depth ambiguities originating from the monocular captures. In addition to guiding the optimization process, we show that these priors can be exploited to explicitly initialize our scene representation to drastically improve optimization speed and final image quality. As evidenced by our experimental evaluation, our dynamic point cloud model not only enables fast optimization and real-time frame rates for interactive applications, but also achieves competitive image quality on monocular benchmark sequences. Our code and data are available online https://moritzkappel.github.io/projects/dnpc/.Computer Graphics ForumFix it in Post: Image and Video Synthesis and Analysis44
A Transparent and Efficient Extension of IceT for Parallel Compositing on Non-Convex Volume Domain Decompositions
No full textThe IceT library is widely used for parallel compositing but does not support non-convex volume domain decompositions. We provide a backward-compatible extension of IceT to handle non-convex domain decompositions of volume data. These are frequently produced in numerical simulations, but it is challenging to render them in parallel due to the non-commutativity of alpha compositing. We enable parallel volume rendering of non-convex domains in IceT by extending its parallel compositing to layered images. Our code follows an embedded design, extending and generalizing IceT's internal functions for image compression, splitting, compositing, and decompression to efficiently handle layered images, while maintaining the existing functionality and API. We perform scalability tests and provide our implementation open-source in a public repository, with in-line documentation and integration tests.Eurographics Symposium on Parallel Graphics and VisualizationPaper
From HBIM to Digital Twins: An Interoperable Framework for Semantic Knowledge Integration and Dynamic Monitoring of Historic Buildings
No full textHistoric Building Information Modeling (HBIM) has established itself as a promising approach for the digital documentation and management of cultural heritage. However, there remain fundamental challenges in terms of semantic depth, interoperability, and the dynamic updatability of such models. This paper proposes a multi-layered framework that integrates a geometry-based HBIM model with ontological knowledge representation, external linked data sources, and sensor-based monitoring. Based on open standards such as IFC, ifcOWL, and CIDOC CRM, a concept for a semantically enriched digital twin is developed, which processes both static information and dynamic environmental and condition data. The framework enables context-based analyses, predictive conservation strategies, and promotes interdisciplinary collaboration. Using a prototype framework, methodological advantages, technological challenges, and future development fields are discussed. The paper thus contributes an integrative impulse to the further development of data-driven, sustainable heritage conservation within the context of digital cultural heritage infrastructures.Digital HeritageFrom 3D Models to Digital Platforms and Digital Twin