- Spring 2018
This is an advanced course on computer vision from a probabilistic and machine learning perspective. It covers techniques such as linear and non-linear filtering, geometry, energy function methods, markov random fields, conditional random fields, graphical models, probabilistic grammars, and deep neural networks. These are illustrated on a set of vision problems ranging from image segmentation, semantic segmentation, depth estimation, object recognition, object parsing, scene parsing, action recognition, and text captioning. [Analysis or Applications] Required course background: calculus, linear algebra (AS.110.201 or equiv.), probability and statistics (AS.553.311 or equiv.), and the ability to program in Python and C++. Background in computer vision (EN.601.461/661) and machine learning (EN.601.475) suggested but not required.