Eurographics Digital Library
Not a member yet
17121 research outputs found
Sort by
Perception of Drawing Reference Quality among Professional Hand-drawn Animators
No full textWe present a preliminary experiment, investigating professional hand-drawn animators' perception of how good one frame is as drawing reference for another. 10 professional hand-drawn animators rated the drawing reference quality of 54 hand-drawn frame pairs, each differing by character pose and region rotation, reflection, and distortion transformations. Our results indicate that animators perceive frames differing by rotation/reflection as better drawing reference than frames differing by distortion.ACM/EG Expressive Symposium - WICED: Eurographics Workshop on Intelligent Cinematography and Editing - Artworks, Posters, DemosPoster
Monitoring Cultural Heritage in the Nagorno-Karabakh Conflict: The Mission and Methods of Cultural Heritage Watch (CHW)
No full textThis paper discusses satellite-based monitoring of cultural heritage sites in the Nagorno-Karabakh region of the South Caucasus, which for decades has been the center of a long-simmering territorial and ethnic conflict between Armenia and Azerbaijan. Following the ceasefire that concluded the Second Nagorno-Karabakh War in fall 2020, the co-authors formed CHW to address the lack of evidence- based documentation of past and present abuses of cultural heritage in this intractable conflict.Digital HeritageDigital Tools for Monitoring Heritage at Ris
A Blender Add-on for 3D Concept Sketching
No full textWe present a Blender add-on that bridges the Grease Pencil drawing interface with a symmetry-driven 3D reconstruction algorithm for concept sketches. On the one hand, Blender's Grease Pencil offers users a convenient interface to draw 2D concept sketches using vector strokes, drawn freehand or using vector primitives (straight lines, Bézier curves). On the other hand, the reconstruction algorithm of Hähnlein et al. [HGSB22] can lift such 2D sketches to 3D using symmetry correspondences. While this reconstruction algorithm was originally developed to be automatic, we show how it can be adapted to allow step-by-step reconstruction of a sketch as it is drawn, and to support user corrections. Finally, we compare the reconstructions obtained with this add-on to the ones produced by a recent generative model trained to produce a 3D shape from a single image.ACM/EG Expressive Symposium - WICED: Eurographics Workshop on Intelligent Cinematography and Editing - Artworks, Posters, DemosPoster
Designing and developing ICH Atlas: A Trend Identification Platform for Digital Valorization of Intangible Cultural Heritage
No full textSocial media-based trend identification platforms powered by Artificial Intelligence (AI) are gaining considerable attention from scholars and industry professionals alike. These platforms analyze large amounts of user-generated data, identify the emergence and evolution of practices, and support data-driven decision-making. Although automatic trend analysis is growing, limited attention has focused on developing platforms that target digital transformation practices in Intangible Cultural Heritage (ICH). This paper introduces the development process behind ICH Atlas, a digital platform that displays and enables dynamic navigation of emerging trends connected to ICH-related professional roles, skills, and digital technologies and their expected growth based on global trends. The platform was the result of nine-month remote collaboration between a university-based design research group with expertise in cultural heritage and a trend forecasting company specializing in artificial intelligence (AI) and big data analytics. Based on a qualitative analysis of meeting records, email correspondence, datasets, and platform prototypes, the paper outlines the platform's iterations. We trace the decision-making process development, encountered challenges, and coping strategies. Based on our reflections, we identify three tensions that might be of interest for industry-academia initiatives in the intangible cultural heritage sector: scaffolding vs ambiguity, interpretation vs granularity of data, tacit vs explicit knowledge.Digital HeritageStandards and Community-Driven Tool
Neural Acquisition & Representation of Subsurface Scattering
No full textWe present a method to acquire and estimate the sub-surface scattering properties of light transport at a highly detailed level by learning the pixel footprint response at each point on the object surface. The reconstruction leverages 3D scanning techniques as input to a U-Net CNN. A stereo projector-camera setup using phase-shifted profilometry (PSP) patterns efficiently captures the data for a variety of scattering objects. Reconstructing dense pixel footprints allows for relighting with arbitrary high-resolution projector patterns. The final output is a relit color image. Qualitative and quantitative comparison against illuminated realworld captured images demonstrate that the predicted footprints are almost identical to the actual responses. The same model is trained for multiple views across multiple objects such that the learned representations can be used to generalize to unseen sub-surface scattering materials as well.Vision, Modeling, and VisualizationNeural and Differentiable Renderin
ER-Diff: A Multi-Scale Exposure Residual-Guided Diffusion Model for Image Exposure Correction
No full textThis paper proposes an Exposure Residual-guided Diffusion Model (ER-Diff) to address the performance limitations of existing image restoration methods in handling non-uniform exposure. Current exposure correction techniques struggle with detail recovery in extreme over/underexposed regions and global exposure balancing. While diffusion models offer powerful generative capabilities for image restoration, effectively leveraging exposure information to guide the denoising process remains underexplored. Additionally, content reconstruction fidelity in severely degraded regions is challenging to ensure. To tackle these issues, ER-Diff explicitly constructs exposure residual features to guide the diffusion process. Specifically, we design a multi-scale exposure residual guidance module that first computes the residual between the input image and an ideally exposed reference, then transforms it into hierarchical feature representations via a multi-scale extraction network, and finally integrates these features progressively into the denoising process. This design enhances feature representation in locally distorted exposure areas while maintaining global exposure consistency. By decoupling content reconstruction and exposure correction, our method achieves more natural exposure adjustment with better detail preservation while ensuring content authenticity. Extensive experiments demonstrate that ER-Diff outperforms state-of-the-art exposure correction methods in both quantitative and qualitative evaluations, particularly in complex lighting conditions, effectively balancing detail retention and exposure correction.Pacific Graphics Conference Papers, Posters, and DemosEnhancing Image
App AskGate - Four Steps in Ascalon
No full textThe development of the app called AskGate originated from the results obtained by the Italian Archaeological Mission MAECI (Italian Ministry of Foreign Affairs and International Cooperation). The research group, consisting of architects, archaeologists, physicists and artists since 2020, has been working on the case study of the so-called church of Santa Maria in Viridis, within the ancient city of Ascalona (Israel) under the direction of Prof. C. M. R. Luschi. The main challenge was to find a language capable of expressing the process of scientific research which would involve the user. The theme of the representation adopted for the app entrusts the narrative in its significant parts to the eloquence of the image, respecting the gestalt linked to the use of the smartphone.Digital HeritagePoster
Acquisition and digitization of large scale heritage scenes with open source project https://github.com/MapsHD/HDMapping
No full textThis paper describes an open source project https://github.com/MapsHD/HDMapping for large-scale 3D mapping using an open-hardware hand-held LiDAR measurement device available at https://github.com/JanuszBedkowski/mandeye_controller in application of large-scale heritage scenes acquisition and digitization. It implements multi view terrestrial laser scanning algorithms, LiDAR odometry, Pose Graph SLAM (Simultaneous Localisation and Mapping), NDT (Normal Distributions Transform) and ICP (Iterative Closest Point). Mobile mapping systems is based on LiVOX MID360 - laser scanner with non repetitive scanning pattern equipped with equirectangular camera. This project runs from 2023, the current version v0.76 has significant improvements in lidar odometry and georeferencing. The goal of the project is to provide an affordable mobile mapping system and an open source software that can be widely used also by Culture Heritage community.Digital HeritageDigitization and Segmentatio
VISLIX: An XAI Framework for Validating Vision Models with Slice Discovery and Analysis
No full textReal-world machine learning models require rigorous evaluation before deployment, especially in safety-critical domains like autonomous driving and surveillance. The evaluation of machine learning models often focuses on data slices, which are subsets of the data that share a set of characteristics. Data slice finding automatically identifies conditions or data subgroups where models underperform, aiding developers in mitigating performance issues. Despite its popularity and effectiveness, data slicing for vision model validation faces several challenges. First, data slicing often needs additional image metadata or visual concepts, and falls short in certain computer vision tasks, such as object detection. Second, understanding data slices is a labor-intensive and mentally demanding process that heavily relies on the expert's domain knowledge. Third, data slicing lacks a human-in-the-loop solution that allows experts to form hypothesis and test them interactively. To overcome these limitations and better support the machine learning operations lifecycle, we introduce VISLIX, a novel visual analytics framework that employs state-of-the-art foundation models to help domain experts analyze slices in computer vision models. Our approach does not require image metadata or visual concepts, automatically generates natural language insights, and allows users to test data slice hypothesis interactively. We evaluate VISLIX with an expert study and three use cases, that demonstrate the effectiveness of our tool in providing comprehensive insights for validating object detection models.Computer Graphics ForumExplainable and Generative A
Beyond Street Signs: Ethical and Situated Cultural Storytelling using AI and Extended Reality
No full textHonorific street names that commemorate historical figures, events, or cultural symbols constitute a frequently overlooked, yet symbolically potent layer of urban heritage. Beyond their practical function in navigation and spatial organization, they also encode ideological narratives into public space that shape collective memory and identity. However, they often go unnoticed, while traditional cultural heritage (CH) tools do not engage with their contested meanings, particularly in politically divided urban landscapes. This paper addresses this gap by exploring how the integration of extended reality (XR) and artificial intelligence (AI) can reanimate toponymic inscriptions into dynamic, context-sensitive forms of storytelling. Focusing on the center of Nicosia, Cyprus, we present an intelligent tourist guide that combines mobile XR, conversational AI, and lifelike avatars to deliver personalized, multilingual narratives, at the sites of honorific street names. We conclude by discussing the ethical challenges of algorithmic memory mediation in contested urban spaces and we argue that such systems, when designed responsibly, can foster critical engagement, pluralistic representation, and new forms of civic pedagogy within the domain of CH.Digital HeritageStorytelling and Interpretation in Digital Heritag