Creation of complex execution and storage mechanisms, based on instruction set architecture, for software design including high-level programming languages and operating systems. Programming design projects. Credit not allowed for both ECE 2035 and ECE 3035.

This course introduces the theory, algorithms, and practical techniques that enable computers to interpret and reason about visual information. Students will study the full pipeline of modern computer vision, beginning with image formation, camera models, and low-level processing (filtering, edge and feature detection, image pyramids), and progressing through mid-level topics such as feature matching, optical flow, image stitching, segmentation, and stereo and multi-view geometry. The latter portion of the course covers high-level vision tasks including object detection, recognition, and tracking, with an emphasis on both classical methods (Kalman filtering, graph cuts, SIFT/ORB) and contemporary deep learning approaches (CNNs, ResNet and transfer learning, R-CNN family detectors, YOLO, and Vision Transformers).

This course teaches system-level design, including both software and hardware. Through activities and projects, students gain exposure to entrepreneurship, product lifecycle management, prototyping, and testing.

The practical application of the finite-difference time-domain and finite element techniques to electromagnetic problems. Computer projects are required.

The design, foundation, and applications of generative and geometric deep learning

This course introduces graduate students to the theory and practice of applying artificial intelligence and machine learning to semiconductor process technology, metrology, manufacturing, and digital twins. The semiconductor industry is undergoing a transformation driven by the increasing complexity of advanced process nodes and the explosion of data generated in modern fabrication facilities. AI/ML techniques and digital twin frameworks are now critical tools for process modeling, equipment monitoring, defect inspection, yield optimization, and accelerating technology development cycles.

Designed for students with a background in semiconductor devices and processes but no prior AI/ML experience, the course begins with a rigorous introduction to machine learning fundamentals using Python, then progressively applies these techniques to real-world semiconductor challenges. Topics include surrogate modeling for TCAD, physics-informed neural networks, virtual metrology, SEM defect classification, yield prediction, fault detection, digital twins for fab modules, and ML-assisted technology development.

The course emphasizes hands-on learning through four Python labs using open semiconductor datasets, two problem-set homeworks, and a substantial semester-long research project. Industry guest lectures provide direct exposure to how these methods are deployed in production fabs and R&D environments.

Placeholder

Placeholder