Hands-on Deep Learning SS 2024

Quick Info

In SS 2024, the lab course will be held twice, i.e., two separate groups.

Time and Place

Block course offered every Winter & Summer Semester
Please sign up using the TUMonline matching system
Group 1: 02.09.-10.09, 9:00-16:30
Group 2: 16.09.-24.09, 9:00-16:30
Place: MW 1050
Regular presence is required

Program

M.Sc. Lab Course (Hochschulpraktikum)
Language: English
4 ECTS, 4 PR

Prerequisits (highly recommended):
- Experience with Python
- A prior course on machine learning

Please deregister from TUMonline if you do not plan to attend or lack the prerequisits to give other students the chance to take the spot.

Links

Module Handbook
Course Calendar
Moodle

Literature

Suggested literature will be given in the course material.

Exam

Type: group project report and presentation
Time and Place: last session

Grade Breakdown

70% project report
30% project presentation

What's this course about?

Get hands-on experience with deep learning algorithms. If you want to learn how autonomous vehicles perceive their environment or how to generate deep fakes, this is the course for you! We will cover deep neural networks, convolutional networks, recurrent nets, autoencoders and deep generative methods. For a more detailed description see the module handbook.

How does it work?

TRICKS of the TRADE

In the 1st hour of each session, the topic is introduced by the TA with slides. Key koncepts are discussed and examples are given from the scientific and every day life.

Learn by doing

In each session, we will learn a different deep learning method by applying it to a concrete problem. Downloadable jupyter notebooks will give you step-by-step instructions to the solution. The solutions will be in Numpy and PyTorch.