Picture for Karthik Duraisamy

Karthik Duraisamy

Spatially-Aware Diffusion Models with Cross-Attention for Global Field Reconstruction with Sparse Observations

Add code
Aug 30, 2024
Figure 1 for Spatially-Aware Diffusion Models with Cross-Attention for Global Field Reconstruction with Sparse Observations
Figure 2 for Spatially-Aware Diffusion Models with Cross-Attention for Global Field Reconstruction with Sparse Observations
Figure 3 for Spatially-Aware Diffusion Models with Cross-Attention for Global Field Reconstruction with Sparse Observations
Figure 4 for Spatially-Aware Diffusion Models with Cross-Attention for Global Field Reconstruction with Sparse Observations
Viaarxiv icon

Enhancing Dynamical System Modeling through Interpretable Machine Learning Augmentations: A Case Study in Cathodic Electrophoretic Deposition

Add code
Jan 16, 2024
Viaarxiv icon

CoCoGen: Physically-Consistent and Conditioned Score-based Generative Models for Forward and Inverse Problems

Add code
Dec 16, 2023
Viaarxiv icon

Easy attention: A simple self-attention mechanism for Transformers

Add code
Aug 24, 2023
Viaarxiv icon

On the lifting and reconstruction of dynamical systems with multiple attractors

Add code
Apr 24, 2023
Viaarxiv icon

Conditionally Parameterized, Discretization-Aware Neural Networks for Mesh-Based Modeling of Physical Systems

Add code
Oct 08, 2021
Figure 1 for Conditionally Parameterized, Discretization-Aware Neural Networks for Mesh-Based Modeling of Physical Systems
Figure 2 for Conditionally Parameterized, Discretization-Aware Neural Networks for Mesh-Based Modeling of Physical Systems
Figure 3 for Conditionally Parameterized, Discretization-Aware Neural Networks for Mesh-Based Modeling of Physical Systems
Figure 4 for Conditionally Parameterized, Discretization-Aware Neural Networks for Mesh-Based Modeling of Physical Systems
Viaarxiv icon

Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries

Add code
Sep 17, 2021
Figure 1 for Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries
Figure 2 for Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries
Figure 3 for Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries
Figure 4 for Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries
Viaarxiv icon

Disentangling Generative Factors of Physical Fields Using Variational Autoencoders

Add code
Sep 15, 2021
Figure 1 for Disentangling Generative Factors of Physical Fields Using Variational Autoencoders
Figure 2 for Disentangling Generative Factors of Physical Fields Using Variational Autoencoders
Figure 3 for Disentangling Generative Factors of Physical Fields Using Variational Autoencoders
Figure 4 for Disentangling Generative Factors of Physical Fields Using Variational Autoencoders
Viaarxiv icon

Sparsity-promoting algorithms for the discovery of informative Koopman invariant subspaces

Add code
Feb 25, 2020
Figure 1 for Sparsity-promoting algorithms for the discovery of informative Koopman invariant subspaces
Figure 2 for Sparsity-promoting algorithms for the discovery of informative Koopman invariant subspaces
Figure 3 for Sparsity-promoting algorithms for the discovery of informative Koopman invariant subspaces
Figure 4 for Sparsity-promoting algorithms for the discovery of informative Koopman invariant subspaces
Viaarxiv icon

Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics

Add code
Dec 23, 2019
Figure 1 for Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics
Figure 2 for Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics
Figure 3 for Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics
Figure 4 for Multi-level Convolutional Autoencoder Networks for Parametric Prediction of Spatio-temporal Dynamics
Viaarxiv icon