Topics in deep learning for perception, 3 credits
Course information
Research education subject
- Computer Science
Course Syllabus
Contacts
-
Todor Stoyanov, Universitetslektor
todor.stoyanov@oru.se -
Josefin Unander-Scharin, Utbildnings- och forskningsadministratör
+46 19 303909 josefin.unander-scharin@oru.se
Course content
This course is intended as a guided study of deep neural network (NN) architectures applied to problems in perception research. We will overview the most common network designs and focus on applications relevant to perception with mobile robots. The course will cover:
- An overview / revision of basic machine learning concepts relevant to deep learning
- Deep feedforward networks
- Autoencoders
- Convolutional neural networks
- Regularization and optimization for NNs
- Practical tips for applying deep learning
- Selected example deep NN approaches for perception problems