Search

Neural radiance fields have emerged as a dominant paradigm for creating complex 3D environments incorporating synthetic novel views. However, 3D object removal applications utilizing neural radiance fields have lagged behind in effectiveness, particularly when open set queries are necessary for determining the relevant objects. One such application area is in architectural heritage preservation, where the automatic removal of non-architectural objects from 3D environments is necessary for many downstream tasks. Furthermore, when modeling occupied buildings, it is crucial for modeling techniques to be privacy preserving by default; this also motivates the removal of non-architectural elements. In this paper, we propose a pipeline for the automatic creation of cleaned, architectural structure only point clouds utilizing a language embedded radiance field (LERF) with a specific application toward generating suitable point clouds for the structural integrity assessment of occupied buildings. We then validated the efficacy of our approach on the rooms of the historic Sion hospital, a national historic monument in Valais, Switzerland. By using our automatic removal pipeline on the point clouds of rooms filled with furniture, we decreased the average earth mover’s distance (EMD) to the ground truth point clouds of the physically emptied rooms by 31 percent. The success of our research points the way toward new paradigms in architectural modeling and cultural preservation.

Neural radiance fields have emerged as a dominant paradigm for creating complex 3D environments incorporating synthetic novel views. However, 3D object removal applications utilizing neural radiance fields have lagged behind in effectiveness, particularly when open set queries are necessary for determining the relevant objects. One such application area is in architectural heritage preservation, where the automatic removal of non-architectural objects from 3D environments is necessary for many downstream tasks. Furthermore, when modeling occupied buildings, it is crucial for modeling techniques to be privacy preserving by default; this also motivates the removal of non-architectural elements. In this paper, we propose a pipeline for the automatic creation of cleaned, architectural structure only point clouds utilizing a language embedded radiance field (LERF) with a specific application toward generating suitable point clouds for the structural integrity assessment of occupied buildings. We then validated the efficacy of our approach on the rooms of the historic Sion hospital, a national historic monument in Valais, Switzerland. By using our automatic removal pipeline on the point clouds of rooms filled with furniture, we decreased the average earth mover's distance (EMD) to the ground truth point clouds of the physically emptied rooms by 31 percent. The success of our research points the way toward new paradigms in architectural modeling and cultural preservation. [ABSTRACT FROM AUTHOR]

Red cultural resources are pivotal in enhancing the pedagogical impact of ideological and political education within higher education institutions. The process of digitally converting these resources significantly broadens the avenues for communication, enriches the content, and elevates the efficiency of dissemination. This investigation delves into the synergistic fusion of virtual reality (VR) technologies with red cultural assets, leveraging the distinct capabilities of VR— namely its immersive, interactive, and perceptual qualities—to augment the efficacy of ideological and political theory instruction in universities. To refine the graphical detail within the virtual environments, this study introduces an adaptive Level of Detail (LOD) model generation algorithm. This novel approach builds upon traditional LOD techniques by incorporating a mesh simplification strategy that utilizes edge collapse. The cost of edge collapse is quantified through a combination of the quadratic error metric and the product of model surface curvature with the length of the collapsed edge. This method facilitates the rapid generation of recursive mesh models, enhancing the visual quality and performance of the VR experience. The practical application of virtual red cultural resources in civics education was evaluated through a comparative study. The empirical findings indicate that the integration of these resources into civics classes resulted in a significant improvement in student performance, with the experimental group outperforming the control group by an average of 11.01 points in post-test assessments. Additionally, the majority of students exposed to the VR-enhanced teaching methodology expressed a high degree of satisfaction, underscoring the potential of virtual cultural resources to enrich traditional educational settings. This demonstrates the transformative potential of integrating advanced digital tools with red cultural content in academic contexts.

This paper establishes different detail-level terrain grids for different terrain data blocks and generates multi-resolution terrain based on the tree-structured LOD model as a way of layering large-scale terrain data. Then, the virtual environment’s visual and image texture are rendered to visualize the virtual spatial environment. Then, the overall architecture of the interaction platform was designed, and the virtual space design visualization and interaction effects were studied based on spatial distance perception, presence experience perception, and error analysis. The results show that there is a significant difference between the mean value of people’s perceived satisfaction in the interaction space and non-interaction space M=163.138, t=1.376, p=0.005

In this study, we successfully realized the interaction design of virtual Civics teaching scene through multiresolution hierarchical detail (LOD) model, and constructed an immersive teaching environment using texture mapping technology. Combined with the gesture recognition and training of Adaboost classifier, this paper proposes an innovative interaction model for virtual Civics and Politics teaching. Through in-depth Analysis of the virtual interactive teaching elements of Civics courses in colleges and universities, this study conducted a covariance analysis of the aspects of creativity, critical thinking, metacognition, self-efficacy, and scientific ideological reasoning. Then it explored the effects of different places of origin and types of schools on the acceptance of virtual Civics courses by the students. It was found that among the science and engineering and teacher training students, the science and engineering students paid more attention to the virtual Civics teaching (3.36 and 3.05, respectively). In addition, compared with traditional Civics teaching, virtual practice teaching significantly increased students’ motivation (3.79), proving that Civics courses incorporating virtual interaction design have significant advantages in improving learning attractiveness.