Image Analysis and Computer Vision II
Prof. Dr. Bastian Goldlücke
|Tue||13:30 - 15:00||G 304|
|Fri||10:00 - 11:30||G 304|
The lecture 'Image Analysis and Computer Vision II' continues from last semester, but is open to students who have not attended part I - all prerequisites will be briefly reviewed. We start with the analysis of image sequences, and learn how to recover motion of image features and track objects in a video. The second part is devoted to a crash course in machine learning, and shows how to implement trainable classifiers which can discriminate between different classes. Vision applications lie for example in optical character, face and object recognition or classification of images or individual pixels. The final part of the lecture will give a brief introduction into the framework of statistical and graphical models, which unifies a lot of problems in image analysis.
Special cases are the popular Markov Random Fields (MRFs) for image segmentation and labeling problems, or Kalman filters for motion models.
The lecture will cover both the modeling perspective, as well as a few selected solvers for inference and training (graph cuts, belief propagation ...).