DEDRECON is an army project which develops advanced machine learning to real-life image visual perception under multi-modality fusion techniques.

Funding: Army Futures Command LOA-2495-018

Invariance-based Multi-Clustering of Latent Space Embeddings for Equivariant Learning

Sample Efficient Learned Decision Making in Controllable Stochastic Dynamical Systems

Video Imputation

Spectre studies smart data structures and prediction optimization with provable guarantees for spectral and geometric processing.

Semi-Supervised Variational Inference for Generative Materials Design

Inverse Generative Modeling for Stealth and Cloaking Devices using Meta-materials

Angstrom studies multiresolution geometric data structures and computational mathematics algorithms that are essential for a wide range of molecular structure determination, energetics, interactions, and simulations.

GEMS data prediction

UG2+ challenge

Tensor Networks for Learning Quantum Variational Latent Spaces

NeuroModeling Understanding the fundamental relationship between neuron structure and function has long been an important goal in neuroscience. At all scales of analysis, the roles that geometric shapes and spatial interrelationships play in determining the functional abilities and constraints on brain activity are of paramount consideration.

Bio-Model studies computational modeling, simulation and visualization of anatomical and physiological processes.

X-Tierra studies computational modeling, simulation and visualization of computational earth sciences.

DiDi studies both the computational hardware and software infrastructure that underpin interactive visual manipulation of extremely large data streams on huge stereo display screens (Data Intensive, Display Intensive computations).