Inhalte

This course addresses state-of-the-art computer vision algorithms that let computers see, learn, and understand image and video content. After being taught the required basics in machine learning, students will - accompanied by practical exercises - get to know the most promising techniques.

The topics of the course may be summarized as follows:

• Machine learning foundations
• Deep learning, with a focus on CNNs and current reference architectures
• Data reduction (quantization, dimensionality reduction)
• Traditional computer vision (hand-crafted features and algorithms)
• CNN-based computer vision

The learned concepts will be illustrated by successful examples in practice. The accompanying exercises will contain some hands-on assignments. Towards the end of the course more advanced topics in object detection and object recognition will be addressed.

Hinweis: Diese Vorlesung ersetzt die frühere Vorlesung „Multimedia II“, kann jedoch genauso eingebracht werden.

Übungen

Es erscheint wöchentlich ein Übungsblatt zu den behandelten Vorlesungsinhalten. Jedes Übungsblatt wird in der Globalübung der folgenden Woche besprochen. Es gibt keine Abgabe / Korrektur von Übungsblättern.

Literatur

• M. Mitchell. Machine Learning. McGraw-Hill Science/Engineering/Math, 1997; Chapters 1-8; ( PDF)
• Ian Goodfellow, Yoshua Bengio, Aaron Courville. Deep Learning. MIT Press, 2016, ISBN-13: 978-
  0262035613; Chapters 2-5 are a must read! ( PDF)
• David A. Forsyth and Jean Ponce. Computer Vision: A Modern Approach. Prentice Hall, Upper Saddle River, New Jersey 07458.( PDF)