Fall 2025: ECGR 4422/ 5122: Random Processing
Course Description: This course provides an introduction to the fundamentals of random variables, random signals, and simulation of random phenomena. The emphasis will be on developing the analysis and design tools needed to apply random process theory to graduate electrical engineering courses and research. Topics include random variables and their key characteristics, sequences of random variables and central limit theorem, properties of random processes, correlation and spectral analysis, linear systems with random inputs, and prediction of random signals. The course will provide examples of applications in queuing network, communications, power load distribution, medical imaging, and robotics.
Spring 2026: ECGR 6090/ 8090: Computational Imaging
In many real-world engineering applications, we are interested in reconstructing an image of interest from measured observations. For example, we may be interested in problems such as estimating a high-quality picture from data acquired with a noisy low-resolution camera in consumer electronics, estimating the 3D anatomic structure of bones from a set of 2D X-ray projections in medical imaging, or imaging the black hole at the center of Galaxy M87 from data from 8 telescopes placed at different location in earth. These are all examples of the application of “computational imaging methods. Computational imaging methods have a wide range of applications in consumer electronics, scientific imaging, HCI, medical imaging, microscopy, and remote sensing.
This course introduces computational imaging methods and their applications to solving inverse imaging problems, including denoising, deconvolution, medical image reconstruction, and parameter estimation. The first part of the course will cover the fundamentals of computational models, regularization-based image reconstruction methods, and classical optimization methods, including gradient-based methods, the proximal method, and the Alternating Direction Method of Multipliers (ADMM). The second part of the course will focus on modern data-driven approaches for solving the inverse problem in imaging, including convolutional neural networks, deep image priors, unrolling networks, Plug-and-Play priors, Regularization by Denoising, Diffusion models, and Physics-informed Neural Networks. The course also includes hands-on Python programming assignments and projects to reinforce theoretical concepts through practical applications.
