Welcome to the Deep Learning for Computer Vision course offered in WS17/18!
Course material
You can access the slides here: here. The password was provided in the lecture.
Tuesdays (18:00-20:00) - 101, Interims Hörsaal 1 (5620.01.101)
Lecturers: Prof. Dr. Laura Leal-Taixé and Prof. Dr. Matthias Niessner.
ECTS: 6
SWS: 4
Tutorial
Thursdays (18:00-19:30) - 00.02.001, MI HS 1, Friedrich L. Bauer Hörsaal (5602.EG.001)
Tutors: Thomas Frerix, Tim Meinhardt, Ji Hou and Andreas Rössler
For details about the exercises, please read the README, which is distributed with the exercise. In particular, you will submit your solutions via this website.
Content
- 17.10 - Lecture 1: Introduction to Computer Vision and history of Deep Learning.
- 19.10 - Lecture 2: Machine Learning basics: linear regression, maximum likelihood.
- 24.10 - Lecture 3: Introduction to neural networks, Back-propagation.
- 07.11 - Lecture 4: Optimization
- 21.11 - Lecture 5: Training Neural Networks Part 1: regularization, activation functions, weight initialization, gradient flow, batch normalization, hyperparameter optimization.
- 28.11 - Lecture 6: Training Neural Networks Part 2: parameter updates, ensembles, dropout.
- 05.12 - Lecture 7: Convolutional Neural Networks.
- 12.12 - Lecture 8: CNN 2
- 19.12 - Lecture 9: CNN 3
- 09.01 - Lecture 10: CNN 4
- 16.01 - Lecture 11: Generative networks, GANs, VAE
- 23.01 - Lecture 12: Recurrent networks and LSTMs
- 30.01 - Guest lecture by Scott Reed, DeepMind
Prerequisites
Passion for mathematics and the use of machine learning in order to solve complex computer vision problems. The course will be focused on practical projects, therefore, previous knowledge of Python is needed.
Moodle
We use Moodle for discussions and to distribute important information. Please check the News and Discussion boards regularly or subscribe to them.
Exercise schedule
EXERCISE 0:
- Topics: Getting familiar with Python
- Starting date: 17.10.2017
- Due date: 01.11.2017
- Recommended link: Python introduction
EXERCISE 1:
- Topics: Linear classifiers, multinomial regression, two-layer neural net.
- Starting date: 02.11.2017
- Due date: 15.11.2017
EXERCISE 2:
- Topics: Fully connected nets, dropout, batch normalization.
- Starting date: 16.11.2017
- Due date: 29.11.2017
EXERCISE 3:
- Topics: Convolutional neural networks, large-scale project with PyTorch.
- Starting date: 30.11.2017
- Due date: 17.12.2017
FINAL PROJECT:
- Topic: Each group will make a project proposal.
- Group matching deadline: 17.12.2017 12pm
- Group matching final done by us: Until 17.12.2017 11.59pm
- Project proposal due date: 20.12.2017 11.59pm
- Project proposal feedback from us: Until 21.12.2017 11.59pm
- Final project start: 22.12.2017
- Poster presentation: 06.02.2018
FINAL EXAM:
- Exam date: 13.02.2018 (16:00-17:30)
If you have any questions regarding the organization of the course, do not hesitate to contact us at: dl4cv@vision.in.tum.de
For questions on the syllabus, exercises or any other questions on the content of the lecture, we will use the Moodle discussion board.
We offer to discuss questions in person in our office hours, in particular regarding the course projects:
Mondays 1-3pm: Chengnan Yu (room 02.09.051)
Tuesdays 1-3pm: Rahul Bohare (room 02.09.051)
Wednesdays 1-3pm: Aysim Toker (room 02.09.051)
Thursdays 1-2pm: Andreas Rössler/Ji Hou (room 02.13.035)
Fridays 1-2pm: Andreas Rössler/Ji Hou (room 02.13.035)
