Development of Deep Vision Systems
Development of Deep Vision Systems
Computer vision systems try to mimic human capabilities of visual perception to support time-consuming and labor-intensive tasks like the recognition, localization, and tracking of critical objects. Nowadays, such systems increasingly rely on methods and tools from the field of machine learning to automatically extract useful information from images that can be utilized for decision support and business automation purposes.
This course provides the necessary fundamentals for the development of modern vision systems based on machine learning. The particular focus is on deep neural networks and their capabilities of automated feature learning. More specifically, we consider different types of network architectures, look at the steps of image labeling and data preparation, discuss crucial hyperparameters and evaluation criteria, and review other related aspects, such as 3D vision, hybrid intelligence, and explainable artificial intelligence.
The course has a strong practical focus. At the beginning of the semester, all fundamentals are provided in lecture sessions and hands-on exercises using the programming language Python. Afterwards, students are encouraged to work (in groups) on real projects to apply the methods and concepts learned during the teaching sessions. The results are presented and discussed at the end of the semester.
Please note that the number of participants for this course is limited. Below you can find further information on the registration process. After a successful course registration, more details on course updates, course materials, technical setup, and other information will be announced in StudOn.
Learning objectives and skills
- understand the challenges for developing vision-based systems,
- understand the basic principles of machine learning and deep neural networks in the realm of image processing,
- explain the general pipeline of computer vision systems based on deep neural networks,
- know about state-of-the-art techniques at the intersection of computer vision and machine learning,
- apply technologies for automated image processing in a practical setting,
- compare and evaluate different system configurations,
- work in groups and present their results together,
- develop skills in collaborative interaction with peers.
Course information
| Courses | Lecture: Development of Deep Vision Systems (2,5 SWS)
Exercise: Development of Deep Vision Systems (2,5 SWS) |
| Credits | 5 ECTS |
| Lecturers | Prof. Dr. Patrick Zschech (Lecture)
Prof. Dr. Mathias Kraus (Exercise) |
| Time | Block course in the first half of the semester.
Lecture: Monday 8:00 – 11:15 (5 sessions à 2 x 90 min) Exercise: Thursday 13:15 – 16:30 (1 intro session à 2 x 90 min) and Thursday 9:45 – 13:00 (4 sessions à 2 x 90 min) For further information see schedule. |
| Location | Due to the COVID-19 pandemic, this semester’s course (Winter 2021/2022) will be held exclusively online via Zoom. Please register for the course via StudOn. |
| Recommended prerequisites | Basic knowledge in data analysis techniques, predictive modelling principles, statistics, and machine learning as taught, for example, in the Bachelor course “Business Analytics: Technologien, Methoden und Konzepte”.
Basic programming skills, preferably in Python. |
| Integration in curriculum | Second or third semester |
| Module compatibility | Master International Information Systems (from 2018/19): Module in the section Information Systems – Data & Knowledge (Elective)
Master International Information Systems (from 2016/17): Module in the section Information Systems – Extension Courses (Elective) Master in Marketing (start before WS17/18): Wahlpflichtbereich Modulgruppe “Methoden” Master in Marketing (start since WS17/18): Wahlpflichtbereich Modulgruppe “Data Science” |
| Method of examination | Project report and presentations, partly in groups |
| Grading procedure | Project report (80%) and presentation (20%) |
| Module frequency | Each semester (without guarantee) |
| Workload | Contact hours: 75 h
Independent study: 75 h |
| Module duration | 1 Semester |
| Teaching and examination language | English |
| (Recommended) reading | All relevant material will be provided during the course. |
Registration
The number of participants is limited and requires a registration via StudOn. The registration period is between September 17 – October 18, 2021. If you register before October 11 there is a higher chance of participation with early bird confirmation. In case of excess demand, participants will be selected by drawing lots. Firm confirmation of registration at the latest until October 25, 2021.
Schedule (lecture, exercise, project seminar)
Please note that the schedule is not yet finalized and we reserve the option for changes.
| CW | Organization / Lecture | Dates / Times | Exercise | Dates / Times |
|---|---|---|---|---|
| Course Registration | September 17 – October 18, 2021 | |||
| 42 | Opening Session |
Mo, October 18, 9:45-11:15
|
||
| 43 | 1) Intro Computer Vision |
Mo, October 25, 8:00-11:15
|
||
| 44 | – | Mo, November 1, Holiday | Intro Anaconda / Colab Python | Thurs, November 4, 13:15-16:30 |
| 45 | 2) Artificial Neural Networks | Mo, November 8, 8:00-11:15 | Numerical Computations | Thurs, November 11, 9:45-13:00 |
| 46 | 3) Convolutional Neural Networks | Mo, November 15, 8:00-11:15 | Intro Deep Learning with Pytorch | Thurs, November 18, 9:45-13:00 |
| 47 | 4) Data Preparation, Labeling, Augmentation + 5) Transfer Learning | Mo, November 22, 8:00-11:15 | Neural Networks for Classification | Thurs, November 25, 9:45-13:00 |
| 48 | 6) Evaluation + 7) Additional Aspects | Mo, November 29, 8:00-11:15 | Neural Networks for Object Detection | Thurs, December 2, 9:45-13:00 |
|
Guest Lecture II
|
tba | |||
| 49 | Project Kick-off |
Mo, December 6, 9:45-11:15
|
||
| 50 | Consultation |
Mo, December 13, 9:45-11:15
|
||
| 51 | Guest Lecture by Andreas Maier (FAU) |
Mo, December 20, 8:15-9:00
|
||
| 51 | Consultation |
Mo, December 20, 9:45-11:15
|
||
| 2 | Consultation |
Mo, January 10, 9:45-11:15
|
||
| 3 | Consultation |
Mo, January 17, 9:45-11:15
|
||
| 4 | Consultation |
Mo, January 24, 9:45-11:15
|
||
| 5 | Guest Lecture by Robotron
+ Consultation |
Mo, January 31, 9:45-11:15
|
||
| 6 | Presentation Preliminary Results | Mo, February 7, 9:00-14:30 | ||
| 10 | Submission Project Seminar |
Fr, March 11, 23:59
|
||
| 11 | Final Presentation |
Mo, March 14, 9:00-14:30
|