EM
E. Milchi
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
2 records found
1
Simulating the infiltration of rainwater into solid materials is of broad interest across multiple fields, such as agriculture, hydrology, urban planning, and landslide prevention, yet most rainwater infiltration models rely on numerical simulation and offer little support for interactive visualization. Recent advances in 3D scene representation, particularly 3D Gaussian Splatting (3DGS), enable high-quality and photorealistic scene captures that promise an inherent integration with particle-based physics simulation engines. However, existing 3DGS-based weather frameworks emphasize appearance modeling and do not address the interaction between rainwater and materials within a 3D scene.
In this thesis, we present a proof of concept tackling the physics-informed simulation of rainwater infiltration into solids represented by 3D Gaussian Splatting. We build on the FeatureSplatting pipeline and extend it from single-object editing to an interactive visualization of rainwater infiltration into solid media. Our approach comprises three components: (1) a preprocessing pipeline that prunes Gaussian artifacts and applies point cloud infilling strategies to transpose a 3DGS representation into particle space, (2) an extension to the MPM solver that models rainwater infiltration as the transfer of fluid between macroscopic porous particles, and (3) functionality updating 3D Gaussian appearance parameters driven by per-particle saturation. We evaluate the method on synthetic 2D and 3D scenarios and show that it reproduces expected physical behavior, as well as analyze in depth the main factors that govern the rate of fluid spread. ...
In this thesis, we present a proof of concept tackling the physics-informed simulation of rainwater infiltration into solids represented by 3D Gaussian Splatting. We build on the FeatureSplatting pipeline and extend it from single-object editing to an interactive visualization of rainwater infiltration into solid media. Our approach comprises three components: (1) a preprocessing pipeline that prunes Gaussian artifacts and applies point cloud infilling strategies to transpose a 3DGS representation into particle space, (2) an extension to the MPM solver that models rainwater infiltration as the transfer of fluid between macroscopic porous particles, and (3) functionality updating 3D Gaussian appearance parameters driven by per-particle saturation. We evaluate the method on synthetic 2D and 3D scenarios and show that it reproduces expected physical behavior, as well as analyze in depth the main factors that govern the rate of fluid spread. ...
Simulating the infiltration of rainwater into solid materials is of broad interest across multiple fields, such as agriculture, hydrology, urban planning, and landslide prevention, yet most rainwater infiltration models rely on numerical simulation and offer little support for interactive visualization. Recent advances in 3D scene representation, particularly 3D Gaussian Splatting (3DGS), enable high-quality and photorealistic scene captures that promise an inherent integration with particle-based physics simulation engines. However, existing 3DGS-based weather frameworks emphasize appearance modeling and do not address the interaction between rainwater and materials within a 3D scene.
In this thesis, we present a proof of concept tackling the physics-informed simulation of rainwater infiltration into solids represented by 3D Gaussian Splatting. We build on the FeatureSplatting pipeline and extend it from single-object editing to an interactive visualization of rainwater infiltration into solid media. Our approach comprises three components: (1) a preprocessing pipeline that prunes Gaussian artifacts and applies point cloud infilling strategies to transpose a 3DGS representation into particle space, (2) an extension to the MPM solver that models rainwater infiltration as the transfer of fluid between macroscopic porous particles, and (3) functionality updating 3D Gaussian appearance parameters driven by per-particle saturation. We evaluate the method on synthetic 2D and 3D scenarios and show that it reproduces expected physical behavior, as well as analyze in depth the main factors that govern the rate of fluid spread.
In this thesis, we present a proof of concept tackling the physics-informed simulation of rainwater infiltration into solids represented by 3D Gaussian Splatting. We build on the FeatureSplatting pipeline and extend it from single-object editing to an interactive visualization of rainwater infiltration into solid media. Our approach comprises three components: (1) a preprocessing pipeline that prunes Gaussian artifacts and applies point cloud infilling strategies to transpose a 3DGS representation into particle space, (2) an extension to the MPM solver that models rainwater infiltration as the transfer of fluid between macroscopic porous particles, and (3) functionality updating 3D Gaussian appearance parameters driven by per-particle saturation. We evaluate the method on synthetic 2D and 3D scenarios and show that it reproduces expected physical behavior, as well as analyze in depth the main factors that govern the rate of fluid spread.
As graph neural networks (GNNs) become more frequently used in the biomedical field, there is a growing need to provide insight into how their predictions are made. An algorithm that does this is GNN-SubNet, developed with the aim of detecting disease subnetworks in protein-protein interaction (PPI) networks. GNN-SubNet makes use of a sampling scheme to generate a global explanation in the form of a node mask which indicates each node's importance for all of the GNN's predictions on a dataset. The aim of this study is to validate GNN-SubNet by comparing it with an alternative approach of obtaining global explanations. Instead of obtaining the node mask via a sampling scheme, multiple (local) explanations are optimized per dataset sample, then the node masks are aggregated by either the mean (Mean Aggregation) or the median value (Median Aggregation) per node.
GNN-SubNet is compared with its two modifications firstly by analyzing which disease subnetworks each algorithm detects, and secondly by leveraging metrics devised to assess explainers for GNNs. The results show that all algorithms detect subnetworks associated with cancer. In terms of the metric scores, Mean Aggregation obtains explanations with the highest fidelity, however no algorithm obtains sparse explanations. The study also indicates that GNN-SubNet obtains variate outcomes over multiple runs, and as such the results may not be reproducible. ...
GNN-SubNet is compared with its two modifications firstly by analyzing which disease subnetworks each algorithm detects, and secondly by leveraging metrics devised to assess explainers for GNNs. The results show that all algorithms detect subnetworks associated with cancer. In terms of the metric scores, Mean Aggregation obtains explanations with the highest fidelity, however no algorithm obtains sparse explanations. The study also indicates that GNN-SubNet obtains variate outcomes over multiple runs, and as such the results may not be reproducible. ...
As graph neural networks (GNNs) become more frequently used in the biomedical field, there is a growing need to provide insight into how their predictions are made. An algorithm that does this is GNN-SubNet, developed with the aim of detecting disease subnetworks in protein-protein interaction (PPI) networks. GNN-SubNet makes use of a sampling scheme to generate a global explanation in the form of a node mask which indicates each node's importance for all of the GNN's predictions on a dataset. The aim of this study is to validate GNN-SubNet by comparing it with an alternative approach of obtaining global explanations. Instead of obtaining the node mask via a sampling scheme, multiple (local) explanations are optimized per dataset sample, then the node masks are aggregated by either the mean (Mean Aggregation) or the median value (Median Aggregation) per node.
GNN-SubNet is compared with its two modifications firstly by analyzing which disease subnetworks each algorithm detects, and secondly by leveraging metrics devised to assess explainers for GNNs. The results show that all algorithms detect subnetworks associated with cancer. In terms of the metric scores, Mean Aggregation obtains explanations with the highest fidelity, however no algorithm obtains sparse explanations. The study also indicates that GNN-SubNet obtains variate outcomes over multiple runs, and as such the results may not be reproducible.
GNN-SubNet is compared with its two modifications firstly by analyzing which disease subnetworks each algorithm detects, and secondly by leveraging metrics devised to assess explainers for GNNs. The results show that all algorithms detect subnetworks associated with cancer. In terms of the metric scores, Mean Aggregation obtains explanations with the highest fidelity, however no algorithm obtains sparse explanations. The study also indicates that GNN-SubNet obtains variate outcomes over multiple runs, and as such the results may not be reproducible.