As immersive media increasingly mediate cultural heritage experiences, a key challenge is how to evaluate content deployed across diverse Extended Reality (XR) platforms. This paper proposes a three-axis evaluation framework - Interpretive, Cognitive/Perceptual, and Affective/Sensory - that synthesizes constructs from museum studies, Human Computer Interaction (HCI), and presence theory into a comparative structure. The framework is applied to "Who Killed Helene Pumpulivaara?", a narrative experience produced in both WebXR and Virtual Reality (VR) formats, with a Spatial Augmented Reality (SAR) version underway. Using user feedback from pilot workshops and AI-assisted design workflows, the study identifies how platform affordances reshape meaning, usability, and emotional impact. The resulting matrix supports transparent, design-grounded evaluation and offers transferable insights for XR storytelling across modalities.
Neural rendering has recently emerged as a powerful technique for generating photorealistic 3D content, enabling high levels of visual fidelity. In parallel, web technologies such as WebGL and WebGPU support platform-independent, in-browser rendering, allowing broad accessibility without the need for platform-specific installations. The convergence of these two advancements opens up new possibilities for delivering immersive, high-fidelity 3D experiences directly through the web. However, achieving such integration remains challenging due to strict real-time performance requirements and limited client-side resources. To address this, we propose a hybrid rendering framework that offloads high-fidelity 3D Gaussian Splatting(3DGS)/4D Gaussian Splatting(4DGS) processing to a server, while delegating lightweight mesh rendering and final image composition to the client via depth-based screen-space fusion. This architecture ensures consistent performance across heterogeneous devices, reduces client-side memory usage, and decouples rendering logic from the client, allowing seamless integration of evolving neural models without frequent re-engineering. Empirical evaluations show that the proposed hybrid architecture reduces the computational burden on client devices while consistently maintaining performance across diverse platforms. These results highlight the potential of our framework as a practical solution for accessible, high-quality neural rendering across diverse web platforms. (The project page is available at: https://streamsplat.pengpark.com/)
In the Architectural Design Process (ADP), the communication of design ideas between professionals and stakeholders is a complex task that currently still uses outdated methods. Despite advances in the use of Computer-Aided Design (CAD) and Building Information Modeling (BIM) tools, products are still presented in 2D on flat screens or paper, leaving too much information for the end user to interpret. The use of Augmented Reality (AR) and Mixed Reality (MR), which allow the overlay and direct interaction of digital objects with the real environment, has been shown to be one of the key factors for the next technological breakthrough in the development of the Architecture, Engineering and Construction (AEC) industry. This research uses the concept of serious gaming to propose a workflow for the development of an AR/MR prototype application for the Microsoft HoloLens2 head-mounted device (HMD) capable of capturing the interactions made in an architectural design project and transferring them for execution in BIM software. The prototype application has been tested on indoor and outdoor case scenarios of office layout design and urban furniture implementation showing feasibility of use, though further possible improvements are considered. The results of a preliminary user experience survey conducted, suggest that the proposed approach can facilitate the communication of design ideas between the architect and the end user, offering a more accessible interpretation of design proposals in comparison to conventional methods. Furthermore, it demonstrates its potential for use at AEC.
We present an early-stage framework for the intelligent evaluation and optimization of remote workspaces using Artificial Intelligence (AI), with a vision toward future integration into Digital Twins and Web3D environments. The proposed system leverages computer vision techniques to identify ergonomic, safety, and comfort-related issues from workspace images, and suggests improvements through an interactive checklist interface. Current components include image-based hazard detection using object recognition models (e.g., YOLO), depth and segmentation models, and a rule-based recommendation engine. We also explore the potential integration of ergonomic assessments through pose estimation (e.g., BlazePose) and environmental parameters such as CO2, light, and noise via smart sensors for future development of a Digital Twin interface. A preliminary case study involving a home office and a coworking space demonstrates the effectiveness of the visual analysis pipeline and recommendation features. While real-time sensor integration and full 3D workspace reconstruction remain future work, our results suggest that AI-based workspace analysis can serve as a foundation for scalable, user-centric ergonomic assessment platforms.
This study explores the application of Additive Manufacturing (AM) for the reconstruction the Etemenanki temple tower in Babylon. The research focuses on creating a 3D model from archaeological findings using Computer-Aided Design (CAD) software. The modeling process resulted in two distinct levels of detail: an abstract model emphasizing geometry and a detailed model incorporating textures and colors derived from historical records through rendering. A Polyjet 3D printer was employed to fabricate physical models. The 3D model represents a digital twin of the 3D printed physical model. The design process involved addressing challenges related to the model’s size, the complexity of replicating brickwork and textures, and the limitations inherent in 3D printing. The model was strategically divided into individual parts, facilitating easier 3D printing and assembly. The results highlight the potential of AM in visualizing archaeological sites and its flexibility in representing diverse interpretations of historical structures. This approach allows for both abstract and detailed visualizations within a single model. Furthermore, the study provides a discussion of production costs and time-frames, offering practical insights into the feasibility of 3D printing for architectural reconstructions. The methodology and solutions developed can be applied to other archaeological projects, enabling the creation of detailed and customizable models for research and presentation purposes.
