Search for: All records

Gaussian splatting methods are emerging as a popular approach for converting multi-view image data into scene representations that allow view synthesis. In particular, there is interest in enabling view synthesis for dynamic scenes using only monocular input data---an ill-posed and challenging problem. The fast pace of work in this area has produced multiple simultaneous papers that claim to work best, which cannot all be true. In this work, we organize, benchmark, and analyze many Gaussian-splatting-based methods, providing apples-to-apples comparisons that prior works have lacked. We use multiple existing datasets and a new instructive synthetic dataset designed to isolate factors that affect reconstruction quality. We systematically categorize Gaussian splatting methods into specific motion representation types and quantify how their differences impact performance. Empirically, we find that their rank order is well-defined in synthetic data, but the complexity of real-world data currently overwhelms the differences. Furthermore, the fast rendering speed of all Gaussian-based methods comes at the cost of brittleness in optimization. We summarize our experiments into a list of findings that can help to further progress in this lively problem setting. 

Abstract The Moon generated a long‐lived core dynamo magnetic field, with intensities at least episodically reaching ∼10–100 μT during the period prior to ∼3.56 Ga. While magnetic anomalies observed within impact basins are likely attributable to the presence of impactor‐added metal, other anomalies such as those associated with lunar swirls are not as conclusively linked to exogenic materials. This has led to the hypothesis that some anomalies may be related to magmatic features such as dikes, sills, and laccoliths. However, basalts returned from the Apollo missions are magnetized too weakly to produce the required magnetization intensities (>0.5 A/m). Here, we test the hypothesis that subsolidus reduction of ilmenite within or adjacent to slowly cooled mafic intrusive bodies could locally enhance metallic FeNi contents within the lunar crust. We find that reduction within hypabyssal dikes with high‐Ti or low‐Ti mare basalt compositions can produce sufficient FeNi grains to carry the minimum >0.5 A/m magnetization intensity inferred for swirls, especially if ambient fields are >10 μT or if fine‐grained Fe‐Ni metals in the pseudo‐single domain grain size range are formed. Therefore, there exists a possibility that certain magnetic anomalies exhibiting various shapes such as linear, swarms, and elliptical patterns may be magmatic in origin. Our study highlights that the domain state of the magnetic carriers is an under‐appreciated factor in controlling a rock's magnetization intensity. The results of this study will help guide interpretations of lunar crustal field data acquired by future rovers that will traverse lunar magnetic anomalies.