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.
The creation of high fidelity synthetic data has long been an important goal in machine learning, particularly in fields like finance where the scarcity of available training and test data make it difficult to utilize many of the deep learning techniques which have proven so powerful in other domains. Despite ample research into different types of synthetic generation techniques, which in recent years have largely focused on generative adversarial networks, there remain key holes in many of the architectures and techniques being utilized. In particular, there are currently no techniques available which can generate multiple financial time series of returns concurrently while capturing the specific statistical properties of financial time series and which incorporate extra information that influences the series, such as macroeconomic factors. We propose the Market Conditional Transformer-Encoder Generative Adversarial Network (MC-TE-GAN), a novel generative adversarial neural network architecture that satisfies the aforementioned challenges. When trained on daily data and compared with the benchmark CoMeTS-GAN model, MC-TE-GAN is able to more accurately capture the relevant univariate statistical properties of financial returns such as linear unpredictability and heteroskedasticity. It is also able to capture the multivariate correlations between concurrently generated series. Furthermore, we demonstrate that when incorporating macroeconomic data as conditioning information MC-TE-GAN works as a compelling historical scenario simulator which is able to create realistic synthetic market crashes or bull runs, even in unseen data.
1 January 2025 in Christian Perspectives on Science and Technology
As the broad scale adoption of Artificial Intelligence (AI) and deep learning systems continues to advance, it is more crucial than ever to understand and implement robust ethical frameworks to guide the usage, development, and societal conceptualisation of these technologies. In this paper, we examine the comparative benefits of the Christian virtue ethics tradition towards the proper deployment of AI and its interaction with related brain-computer interface technology. Furthermore, we propose a virtue ethics-informed training recipe for large language models based on the paradigm of reinforcement learning from AI feedback (RLAIF). Lastly, we examine the risk for individuals and society when interfacing with these tools and their impact upon human virtue.
Traditional 3D scene understanding techniques are generally predicated on hand-annotated label sets, but in recent years a new class of open-vocabulary 3D scene understanding techniques has emerged. Despite the success of this paradigm on small scenes, existing approaches cannot scale efficiently to city-scale 3D datasets. In this paper, we present Hierarchical vocab-Agnostic Expert Clustering (HAEC), after the latin word for 'these', a superpoint graph clustering based approach which utilizes a novel mixture of experts graph transformer for its backbone. We administer this highly scalable approach to the first application of open-vocabulary scene understanding on the SensatUrban city-scale dataset. We also demonstrate a synthetic labeling pipeline which is derived entirely from the raw point clouds with no hand-annotation. Our technique can help unlock complex operations on dense urban 3D scenes and open a new path forward in the processing of digital twins.