Colin N Reinhardt

BIO

Dr. Colin N. Reinhardt received a B.S. in Physics and B.S.E.E from University of Washington (2005). DoD/ASEE SMART Doctoral Fellowship recipient (2006-2010), PhD in Electrical Engineering from University of Washington (2010). Dr. Reinhardt has been awarded four US patents, and has more than 30 technical publications. Dr. Colin N Reinhardt joined the Naval Information Warfare Center, Pacific (NIWC-PAC), Communications & Networking Department, in 2011. Dr. Colin N Reinhardt was appointed as affiliate assistant professor in the Department of Electrical & Computer Engineering at the University of Washington, Seattle, in 2017. He currently remains active in both positions.

Dr. Reinhardt's work with the US Navy includes the role as Chief Scientist/Principal Engineer and Lead PI for the TrueView EO/IR physics-based 3D visualization and scene simulation system, a C/C++-based software framework and API which leverages a combination of classical computational/numerical physics models and modern deep learning networks, along with GPU-accelerated massively-parallelized scientific and numerical algorithms and distributed parallel computing on HPC supercomputer architectures. His work involves the application of classical engineered and artificial intelligence (AI) machine/deep learning methods in the areas of EO/IR imagery, computer vision, imaging system performance modeling and prediction, electromagnetic wave propagation theory and modeling through the atmosphere and ocean including turbulence, multiple-scattering, and refractivity phenomena, as well electromagnetic and laser propagation and transmission theory and modeling through random media in various states/phases (solid/liquid/gaseous).

Dr. Reinhardt's teaching and academic work/research is focused on GPU hardware architectures and GPU-accelerated parallel computing, high-performance computing (HPC) systems, parallel processing and parallel software algorithm design/development/testing/optimization, and interactive scientific visualization techniques, frameworks, and applications to complex multi-dimensional spatio-temporally varying massive datasets. 3D computer graphics methods and techniques for rendering and computing using Vulkan/OpenGL APIs andc graphics/compute shader pipelines on multiple HW vendor platforms. GPGPU computing with Intel GPU architectures with OpenCL, NVIDIA GPU architectures w CUDA, particularly with application to computationally-intensive scientific solvers and computer vision algorithms.

Dr. Reinhardt also is a US appointee to several NATO STO task groups related to EO/IR signature modeling, EO/IR/HS decision aid systems, and related electromagnetic wave propagation effects, and high fidelity EO/IR/HS physics-based scene simulation.

Current and Selected Research Highlights

Toward quantifying the real-versus-synthetic imagery data 'reality gap': analysis and practical applications
Colin N. Reinhardt, Sarah Brockman, Rusty Blue, Brian Clipp, and Anthony Hoogs
Proc of SPIE. Vol. 13035, Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Techniques, and Applications II
2024

Preliminary results for a Deep Learning Atmospheric Radiative Transfer Emulator to accelerate predictive 3D scene simulation
Colin N Reinhardt, Derek Opel, Jeffrey Clark, Susan Shoemaker, Shibin Paramswaran
Naval Applications of Machine Learning conference, San Diego, CA
2024

Curved Path Ray Marching and Ray-Object Intersection
Reagan Burke, Charles Timme, Jerry Tessendorf, Colin N Reinhardt
PREPRINT (on request)

Enhanced physics-based modeling of the electro-optical/infrared observation chain
Colin N Reinhardt, Zachariah Ritsema, Jeffrey Clark, Derek Opel, Brian Young, Brian Teaney
Laser Communication and Propagation through the Atmosphere and Oceans V, Proc of SPIE Vol. 12106
2022

Visualization of simulated ocean scenes employing whitecap fraction phenomenology
Jerry Tessendorf, Liang Gao, Colin N Reinhardt
Ocean Sciences Meeting 2022
2022

Enhanced Sea-Surface Spectral Radiance Background Signature Modeling and Simulation
Colin N Reinhardt, James A Ritcey
Military Sensing Symposium, BSD,
2016

Efficient physics-based predictive 3D modeling and image simulation of optical atmospheric refraction phenomena
Colin N Reinhardt, Stephen M Hammel, Dimitris Tsintikidis
Laser Communication and Propagation through the Atmosphere and Oceans V, Vol 9979
2016

Machine Learning Methods for the Analysis of Turbulence-and Extinction-Degraded Imagery
Colin N Reinhardt, David T. Wayne, Kevin M. McBryde, Ana K. Ascencio
Imaging and Applied Optics conference, Optical Society of America (OSA)
2014
• This just to show we've been interested in and using imagery-based (CV) machine learning for a while, pre-CNN era.

Improving bit-error-rate performance of the free-space optical communications system with channel estimation based on radiative transfer theory
Colin N Reinhardt, Yasuo Kuga, Sermsak Jaruwatanadilok, and Akira Ishimaru
IEEE Journal on selected areas in Communications 27(9)
2009
• I was privileged and honored to work with such a knowledgeable and experienced group of PhD dissertation commmittee members/advisors. It was a life-/career-shaping experience! Go Huskies!

Current Teaching

GPU-Accelerated Computing & Visualization track, UW ECE, 2024-25 Academic year
• Autumn 2024 : Applied High Performance GPU Computing
• Winter 2025 : GPU-Accelerated Interactive Scientific Visualization Techniques (SciVis)
• Spring 2025 : Advanced GPU Computing and Visualization

CV/RESUME