VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.
Last Updated: 2026-02-05 16:37:18
Abstract
Learning robots presents both significant research challenges and great commercial opportunities. This course explores the research frontiers of robot learning and dives into building practical systems such as autonomous driving and humanoid robots. The lectures will cover advanced topics on machine learning, perception, control, planning, prediction, mapping, and reinforcement learning.
Objective
Students will learn the advanced topics in perception and robotics to understand research frontiers and engineering practices in building learning robot systems. The lectures will cover the foundations in robot learning systems, including dynamic scene understanding, high-level reasoning, and decision making. Despite the immense scopes of those areas, we will focus on the advanced topics directly related to robot learning. The course will equip the students with knowledge and experience to start research works immediately in those areas. At the same time, students will learn how to apply those ideas and methods in practical systems and applications. So those interested in engineering careers can understand the boundaries between research explorations and practical solutions and how real-world robot systems work behind the scene. Students will have a solid grasp of the main ideas and theories for robot learning. Besides, through a series of projects, students will gain hands-on experience building and running state-of-the-art models in dynamic scene understanding and reinforcement learning. Also, students will learn how to experiment with their robot systems in simulation environments.
Content
The course assumes you have taken lectures in computer vision and machine learning, and you are familiar with conducting deep learning experiments. We aim to cover advanced CV and ML topics closely related to robot learning, and get you prepared for research study and advanced engineering solutions. We will cover the following areas and topics: 1) Dynamic 3D scene perception - 2D and 3D object detection and tracking - Multi-task learning - Geometry Processing - Visual localization - Visual mapping 2) Learning and reasoning - Meta-learning - Few-shot learning - Domain adaptation - Interactive learning - Causal reasoning - Lifelong learning 3) Decision making - Imitation learning - Model-free reinforcement learning - Model-based reinforcement learning - Inverse reinforcement learning - Hierarchical reinforcement learning - Learning to predict - Learning to plan 4) Applications - Autonomous driving - Object grasping - Object manipulation - Autonomous exploration The deadline to deregister is two weeks after the first lecture.
Resources
Literature
The course doesn't use a particular textbook, but each lecture will have a reading list.
General Information
- Language
- English
- Levels
- MSC
- Frequency
- Yearly recurring
Examination
- Type
- graded semester performance
Registration & Places
- Max Places
- 32
Course Components
| Type | Title | Time & Place | Hours |
|---|---|---|---|
| lecture | Robot Learning |
|
2 h weekly |
| exercise | Robot Learning |
|
2 h weekly |
Offered In
-
-
-
Track: Signal Processing and Machine Learning (The core courses and specialization courses below are a selection for students who wish to specialize in the area of "Signal Processing and Machine Learning ", see . The individual study plan is subject to the tutor's approval.)
-
Specialization Courses (These specialization courses are particularly recommended for the area of "Signal Processing and Machine Learning", but you are free to choose courses from any other field in agreement with your tutor. A minimum of 40 credits must be obtained from specialization courses during the MSc EEIT.)
-
-
-
-
Major Courses (A total of 42 CP must be achieved form courses during the Master Program. The individual study plan is subject to the tutor's approval.)
-
-
-