Digital twins have become a central paradigm for cultural heritage documentation, monitoring, and preventive preservation. Yet, when cultural heritage systems promise prediction, simulation, intervention planning, and decision support, a more explicit account is needed of the computational commitments behind such claims. This position paper proposes the notion of the heritage world model as a conceptual and architectural abstraction that uses the semantic digital twin as its representational layer and extends it toward prediction, memory, uncertainty-aware reasoning, and intervention evaluation. We define a heritage world model as a structured, temporally updated, semantically grounded, and action-aware model of a heritage asset and its preservation environment, capable of integrating observations, estimating latent risk states, predicting plausible future trajectories, and evaluating interventions under uncertainty. The paper does not present a validated deployed system. Rather, it clarifies the architectural conditions under which a decision-support digital twin infrastructure could support the kind of world-model-like preservation system proposed here. It further argues that such a model becomes operationally meaningful only when it includes a human-supervised controller layer that maps semantic state, predicted risk trajectories, uncertainty, memory, and institutional constraints into preservation-relevant actions, alerts, monitoring adaptations, or requests for expert review. Sensor data, remote sensing, computational models, risk assessments, policies, and conservation actions are interpreted as possible observational, dynamic, and intervention layers of a heritage world model. The paper reviews adjacent work in heritage digital twins, semantic and reactive ontologies, risk-aware preservation, agentic AI, and modern AI world models, and proposes a research agenda for moving toward predictive, memory-bearing, and intervention-aware preservation intelligence.

Archaeologists, as well as specialists and practitioners in cultural heritage, require applications with additional functions, such as the annotation and attachment of metadata to specific regions of the 3D digital artifacts, to go beyond the simplistic three-dimensional (3D) visualization. Different strategies addressed this issue, most of which are excellent in their particular area of application, but their capacity is limited to their design's purpose; they lack generalization and interoperability. This paper introduces ART3mis, a general-purpose, user-friendly, feature-rich, interactive web-based textual annotation tool for 3D objects. Moreover, it enables the communication, distribution, and reuse of information as it complies with the W3C Web Annotation Data Model. It is primarily designed to help cultural heritage conservators, restorers, and curators who lack technical expertise in 3D imaging and graphics, handle, segment, and annotate 3D digital replicas of artifacts with ease.

Developing language treebanks is a challenging and time-consuming process, especially for under-resourced languages with limited data. This study proposes the use of active learning approaches to automate aspects of this process, aiming to reduce both annotation duration and cost. We propose practical active annotation schemes in which experts strategically select sentences for the training set, initiating a circular process involving annotation prediction, expert correction, and model retraining. To validate the feasibility of these schemes, we applied them to 300 annotated sentences from the newly created Pomak corpus (an under-resourced language), recently published in Universal Dependencies treebanks, as the result of our efforts. Through experiments involving a simple weighted summation of annotation errors, we identified an optimal strategy. This strategy resulted in a 69

Monocular Depth Estimation (MDE) infers per-pixel scene geometry from a single RGB image. Despite recent progress, global MDE models often blur depth discontinuities at object boundaries and fail to capture object-level structure. Segment-aware depth estimation addresses this limitation by exploiting semantic segmentation to decompose depth prediction into simpler, class-specific subproblems. In this work, we study semantic-aware MDE in a multi-branch design where each semantic class is handled by a lightweight Vision Transformer (ViT) branch that predicts dense depth for its class while suppressing interference from other regions. We further examine fusion strategies that merge the branch outputs into a single prediction: (i) a learnable cross-attention fusion module that predicts depth from the stack of per-class proposals and masks, and (ii) a parameter-free stitched summation that sums mask-gated outputs. The proposed architecture is simple, scalable, end-to-end trainable, and compatible with arbitrary transformer backbones. Experiments on Virtual KITTI 2, where ground-truth depth and semantic labels are available, show that segment-aware modeling produces sharper depth boundaries and improves standard error metrics compared to a single-branch baseline (AbsRel 0.243→0.152; RMSE 11.952→9.101). Finally, we find that the parameter-free summation matches, and in most cases improves upon, the accuracy of learned fusion while adding no computational overhead.

The integration of advanced game engines and web technologies has transformed the intersection of culture, education, and gaming. This paper introduces Museatopia, an innovative virtual world framework designed specifically to become the Metaverse for virtual museums and exhibitions. Museatopia supports natural, user-friendly, immersive interaction, allowing, on one hand, the museum owners to easily curate their virtual presence with custom virtual exhibitions, and, on the other hand, the visitors to participate in rich cultural experiences. By combining digital content from various museums, Museatopia provides a dynamic cultural Metaverse, where even small museums can showcase and promote their content on a global scale without any digital skill requirements, greatly supporting their sustainability.

In cultural heritage, various methods have addressed three-dimensional (3D) digitization and reconstruction using two-dimensional (2D) images, including shape from structured light, shape from stereo, and structure from motion. A common challenge is the recognition of the distance of objects in a scene from the viewpoint, usually called depth. Recognizing depth in 2D photographs remains a challenge in computer vision. Deep Learning techniques have improved this area, particularly in what is called monocular depth estimation. This work investigates the applicability of state-of-the-art monocular depth estimation methods as a preliminary step toward complete and automated 3D reconstruction of cultural heritage artifacts. As image-based 3D reconstruction is extremely time-consuming and expertise-demanding, monocular depth estimation methods could provide significant advantages in speeding up the overall process. Our preliminary experiments yield promising results, with a mean accuracy of 88.15

Monocular depth estimation is a central problem in computer vision with applications in robotics, augmented reality, and autonomous driving, yet the self-attention mechanisms used by modern Transformer architectures remain opaque. In this work, we integrate SVD-Inspired Attention (SVDA) into the Dense Prediction Transformer (DPT), introducing a spectrally structured attention formulation for dense prediction that decouples directional alignment from spectral modulation through a learnable diagonal matrix embedded in normalized query–key interactions. Experiments on KITTI and NYU-v2 show that SVDA preserves competitive predictive performance while enabling intrinsic interpretability: on KITTI, AbsRel improves from 0.058 to 0.056 and δ1 from 0.976 to 0.979, while on NYU-v2, AbsRel improves from 0.133 to 0.124 and δ1 from 0.865 to 0.872. This is achieved with only 0.01% additional parameters, at the cost of a measurable runtime overhead associated with the added normalization and spectral modulation. More importantly, SVDA enables six spectral indicators that quantify entropy, rank, sparsity, alignment, selectivity, and robustness, revealing consistent cross-dataset and depth-wise patterns in how attention organizes during training. These properties make the model easier to inspect and better suited to applications where transparency and reliability are important, such as robotics and autonomous navigation.

Vision Transformers (ViTs) have achieved state-of-the-art performance in image classification, yet their attention mechanisms often remain opaque and exhibit dense, non-structured behaviors. In this work, we adapt our previously proposed SVD-Inspired Attention (SVDA) mechanism to the ViT architecture, introducing a geometrically grounded formulation that enhances interpretability, sparsity, and spectral structure. We apply the use of interpretability indicators—originally proposed with SVDA—to monitor attention dynamics during training and assess structural properties of the learned representations. Experimental evaluations on four widely used benchmarks—CIFAR-10, FashionMNIST, CIFAR-100, and ImageNet-100—together with an additional pretrained fine-tuning study in a standard ViT setting show that SVDA preserves competitive classification behavior in our experimental settings while providing descriptive diagnostics of attention structure. In the pretrained setting, we integrate the exact SVDA operator into the late transformer blocks of a standard pretrained ViT and fine-tune on ImageNet-100, providing additional evidence that the proposed mechanism remains viable beyond compact from-scratch training. While the current framework offers descriptive insights rather than prescriptive guidance, our results establish SVDA as a comprehensive and informative tool for analyzing and developing structured attention models in computer vision. This work lays the foundation for future advances in explainable AI, spectral diagnostics, and attention-based model design.

The use of emerging technologies and the creation of open-source textual resources is a significant step towards a thorough linguistic study of the language of the Pomaks. “Philotis” research project (2020–2023) focused on the creation of the infrastructure, the methodology and the tools for the recording, analysis and documentation of living languages, and adopted Pomak as a case study. The corpus of Pomak compiled by “Philotis” contains about 140,000 words selected from a variety of text types including web articles, pieces of news, novels, poems, proverbs, traditional songs, folk tales and language lessons. The resources that resulted from this effort are openly available to Natural Language Processing practitioners and could be helpful for an advanced quantitative and qualitative analysis of the use and structure of Pomak.

Emotion recognition in Virtual Reality (VR) merges technology with humanities research, providing insights into human emotions in immersive environments. With VR expanding into education, healthcare, and social sciences, accurate emotion recognition is essential. This paper addresses the challenge of limited data by proposing a modified MixUp algorithm for VR emotion recognition. Our technique, which selectively combines data points within the same class, improves model generalization and accuracy. Testing on the VREED dataset demonstrates significant enhancements. This study reviews VR and emotion recognition technologies, details the class-intrinsic MixUp method, and discusses implications and future directions for humanities research using VR.

Beyond simplistic 3D visualisations, archaeologists, as well as cultural heritage experts and practitioners, need applications with advanced functionalities. Such as the annotation and attachment of metadata onto particular regions of the 3D digital objects. Various approaches have been presented to tackle this challenge, most of which achieve excellent results in the domain of their application. However, they are often confined to that specific domain and particular problem. In this paper, we present ART3mis - a general-purpose, user-friendly, interactive textual annotation tool for 3D objects. Primarily attuned to aid cultural heritage conservators, restorers and curators with no technical skills in 3D imaging and graphics, the tool allows for the easy handling, segmenting and annotating of 3D digital replicas of artefacts. ART3mis applies a user-driven, direct-on-surface approach. It can handle detailed 3D cultural objects in real-time and store textual annotations for multiple complex regions in JSON data format.

This paper presents the design, development, and pilot deployment of Vitalia, an AR-gamified application targeting the formation of healthy habits in primary education children. Developed within the EU DUSE project, Vitalia integrates physical activity, nutritional education, and immersive storytelling into a gamified framework to promote sustained behavioral change. Grounded in evidence-based behavior change models and co-designed with health, nutrition, and physical activity experts, the system envisions high daily engagement rates and measurable knowledge improvements. The concept positions Vitalia as a scalable model for child-centric, ethically responsible digital health interventions, with the potential to be integrated into school curricula and public health strategies.

In this paper we present PANOPTES, a digital twin ontology designed to support dynamic monitoring, predictive analytics, and decision-making in the management of cultural heritage assets. Building on standards such as CIDOC CRM, SOSA/SSN, PROV-O, GeoSPARQL, and OWL-Time, PANOPTES introduces a unified semantic model that represents assets, observations, diagnoses, predictions, threats, and decision processes. The ontology naturally maps to a relational backend, enabling efficient data ingestion, traceability, and operational deployment. Application scenarios include condition monitoring, risk forecasting, emergency response, and preventive conservation planning. PANOPTES enables structured monitoring, predictive threat modeling, and decision support for the preventive conservation of cultural heritage assets, particularly in remote and infrastructure-poor sites.

The estimation of depth in two-dimensional images has long been a challenging and extensively studied subject in computer vision. Recently, significant progress has been made with the emergence of Deep Learning-based approaches, which have proven highly successful. This paper focuses on the explainability in monocular depth estimation methods, in terms of how humans perceive depth. This preliminary study emphasizes on one of the most significant visual cues, the relative size, which is prominent in almost all viewed images. We designed a specific experiment to mimic the experiments in humans and have tested state-of-the-art methods to indirectly assess the explainability in the context defined. In addition, we observed that measuring the accuracy required further attention and a particular approach is proposed to this end. The results show that a mean accuracy of around 77% across methods is achieved, with some of the methods performing markedly better, thus, indirectly revealing their corresponding potential to uncover monocular depth cues, like relative size.

Self-attention is a cornerstone of modern deep learning, yet its dense dot-product formulation offers limited interpretability and lacks explicit structural constraints. We propose SVD-inspired Attention (SVDA), a novel self-attention mechanism that introduces normalized query/key projections and a learnable diagonal spectral modulation, drawing direct motivation from the structure of Singular Value Decomposition (SVD). This formulation separates directional alignment from spectral emphasis, offering a geometrically grounded and interpretable variant of attention. We formalize SVDA within a standard multi-head Transformer architecture and introduce a suite of structure-aware indicators—such as spectral entropy, effective rank, and selectivity—that quantify interpretability and sparsity in attention dynamics. Our analysis highlights SVDA’s capacity for structured, energy-aware attention without compromising architectural compatibility or expressiveness. This work provides a theoretical foundation and diagnostic framework for structured attention models aimed at interpretability, compression, and semantic transparency.

Remote built heritage sites face increasingly complex threats from climate, environment, and human activity, challenging traditional preservation methods. The ARGUS EU project introduces a novel, sensor-driven strategy for preventive preservation, tailored to remote and infrastructure-poor heritage contexts. This paper presents the threat-to-sensor mapping framework developed in ARGUS, aligning sensor types with site-specific preservation risks. By combining heterogeneous sensing modalities, low-power communication protocols, and risk-informed design, the system supports actionable early warnings and long-term monitoring. Deployments across five pilot sites address key preservation threats: structural deformation, freeze–thaw cycles, environmental degradation, and biological growth. We describe the sensor selection rationale, installation strategies, and early operational insights across varied heritage contexts.

This paper introduces the concept of evolving actor–network ontologies (EANO) as a new paradigm for cultural digital twins. Building on actor–network theory, EANO reframes ontologies from static representations into reflexive, dynamic structures in which semantic interpretations are continuously negotiated among heterogeneous actors. We propose a five-layer architecture that operationalizes this principle, embedding reflexivity, actor salience, and systemic parameters such as resistance and volatility directly into the ontological model. To illustrate this approach, we present minimal simulations that demonstrate how different actor constellations and systemic conditions lead to distinct patterns of semantic evolution, ranging from expert erosion to contested equilibria and balanced coexistence. Rather than serving as predictive models, these simulations exemplify how EANO captures semantic plurality and contestation within a transparent and interpretable framework. The contribution of this work is thus twofold: it provides a conceptual foundation for evolving ontologies in digital heritage and a lightweight demonstration of how such models can be instantiated and explored computationally.

Estimation of depth in two-dimensional images is among the challenging topics in Computer Vision. This is a well-studied but also an ill-posed problem, which has long been the focus of intense research. This paper is an in-depth review of the topic, presenting two aspects, one that considers the mechanisms of human depth perception, and another that includes the various Deep Learning approaches. The methods are presented in a compact and structured way that outlines the topic and categorizes the approaches according to the line of research followed in the recent decade. Although there has been significant advancement in the topic, it was without any connection with human depth perception and the potential benefits from this sector.

Abstract Small cultural institutions play a vital role in preserving and promoting cultural heritage. To ensure their long-term viability and growth, these institutions must undergo a digital transformation. However, they often face significant challenges due to limited resources and a lack of technical expertise. This paper delves into the crucial role of digital tools and services in supporting these small cultural stakeholders. The presentation focuses on project “Thalia”, which embraces this concept through a comprehensive framework that combines modern asset management and innovative marketing strategies with cutting-edge technology. It leverages the CIDOC-CRM data model and a shared web structure to provide secure asset management, virtual exhibitions, digital tours, and connections with crowdfunding portals. Furthermore, a collaborative augmented reality tool fosters stakeholder engagement. Keywords: Cultural Sustainability, Digital Transformation, Small Cultural Institutions, Project Thalia