18-692 Applied Computer Vision

12 units, Carnegie Mellon University

This course provides a comprehensive, applied introduction to computer vision, emphasizing the principles that enable machines to perceive, interpret, and reason about visual data in the physical world. The course begins with the mathematical and geometric foundations of vision, including homogeneous coordinate systems and image formation, and progresses through digital image processing and classical feature extraction methods that shaped early vision systems. Building on these foundations, students are introduced to machine learning approaches for vision, followed by modern deep learning techniques such as convolutional neural networks and transformer-based models. The latter part of the course explores 3D vision, representation learning, and perception tasks, culminating in embodied perception and real-world applications. Throughout the course, concepts are motivated by practical considerations and system-level thinking, preparing students to design, analyze, and implement robust computer vision pipelines.

Prerequisites: none listed

CMU Courses needs JavaScript to show sections, schedules and planning tools.