Display language
To modulepage Generate PDF

#41089 / #1

Seit SoSe 2023

English

Recent Trends in Deep Learning for Computer Vision

3

Demir, Begüm

benotet

Portfolioprüfung

Zugehörigkeit


Fakultät IV

Institut für Technische Informatik und Mikroelektronik

34342200 FG Remote Sensing Image Analysis

No information

Kontakt


EN 5

Burgert, Tom Oswald

t.burgert@tu-berlin.de

Learning Outcomes

Participants of this seminar acquire advanced knowledge in the field of deep learning for computer vision. After completing this course, the students are capable of conducting a literature search, contextualizing scientific papers and presenting topics drawn from the deep learning and computer vision literature. Moreover, students participate in scientific discussions and have an opportunity to sharpen their critical thinking skills.

Content

The seminar aims to deepen the understanding of the participants in current research problems at the intersection of deep learning and computer vision. The topics include, yet are not limited to: - The evolution of CNN architectures (depth, layer composition, activation functions, drop out); - Vision transformers (attention mechanism, basic and advanced architectures); - Data augmentation techniques (basic techniques, MixUp, CutOut, AutoAug, RandAug, etc.); - Semi-supervised learning (temporal ensembling, mean teacher, co-teaching, consistency loss and pseudo-labeling, MixMatch, ReMixMatch, etc.); - Self-supervised learning (foundations of representation learning, pre-text tasks, contrastive loss, triplet loss, SimCLR, BYOL, SimSiam, etc.); - Diffusion Models and Dall-E 2 (current breakthroughs in fusion of semantics of text and image, generative modeling); - Explainable AI (clever hans phenomena, backprogation based method, gradient-based methods); - Optimization theory of neural networks; - Learning and generalization theory of neural networks (shortcut learning, forgetting/memorization events, layer importance, sample difficulty, learning paths, learning bias in texture/shape).

Module Components

Pflichtgruppe:

All Courses are mandatory.

Course NameTypeNumberCycleLanguageSWSVZ
Recent Trends in Deep Learning for Computer VisionSEMSoSeEnglish2

Workload and Credit Points

Recent Trends in Deep Learning for Computer Vision (SEM):

Workload descriptionMultiplierHoursTotal
Attendance15.02.0h30.0h
Pre/post processing15.04.0h60.0h
90.0h(~3 LP)
The Workload of the module sums up to 90.0 Hours. Therefore the module contains 3 Credits.

Description of Teaching and Learning Methods

At the beginning of the semester, there will be an overview lecture that will serve as an introduction to the current research problems in deep learning and computer vision. In addition, a set of primary literature works will be provided to assist the students to choose their topic. A follow-up lecture will guide the students through the process of how to work with scientific papers and write a scientific report. Then, the students will independently investigate the topics assigned to them and prepare a presentation. For the remaining part of the semester, each class will cover one of the course topics. Two students assigned to one topic will give a presentation of approximately 40 minutes (20 minutes each) that will be followed by a group discussion with all participants of the seminar. In addition to the talk, the students will also prepare a scientific report on their topic of approximately 10-15 pages.

Requirements for participation and examination

Desirable prerequisites for participation in the courses:

Good knowledge in Mathematics, especially linear algebra and statistics, are required. The courses Machine Learning I and Machine Learning II or an equivalent should have been completed. Further, the students should have a basic understanding of deep learning which they gained through additional lectures, projects, or seminars. It is recommended to have worked with scientific literature before. This course is particularly designed for students that are in the last semester before writing their master thesis and have a strong interest in deep learning and computer vision.

Mandatory requirements for the module test application:

This module has no requirements.

Module completion

Grading

graded

Type of exam

Portfolio examination

Type of portfolio examination

100 Punkte insgesamt

Language

English

Test elements

NamePointsCategorieDuration/Extent
(Deliverable assessment) Seminar Presentation50oralapprox. 20 minutes
(Deliverable assessment) Written Seminar Report50written10-15 pages

Grading scale

Notenschlüssel »Notenschlüssel 2: Fak IV (2)«

Gesamtpunktzahl1.01.31.72.02.32.73.03.33.74.0
100.0pt95.0pt90.0pt85.0pt80.0pt75.0pt70.0pt65.0pt60.0pt55.0pt50.0pt

Test description (Module completion)

There are are two portfolio deliverable assessments: 1) a long seminar presentation including a discussion (50 points) and 2) a written seminar report (50 points).

Duration of the Module

The following number of semesters is estimated for taking and completing the module:
1 Semester.

This module may be commenced in the following semesters:
Sommersemester.

Maximum Number of Participants

The maximum capacity of students is 18.

Registration Procedures

Students intending to take this seminar need to follow the instructions on the ISIS course page for pre-semester application. Within the first 6 weeks after the commencement of the seminar, students will have to register for the module at Moses (MTS) and additionally at ISIS (TU Berlin's variant of 'Moodle') for teaching materials and communication.

Recommended reading, Lecture notes

Lecture notes

Availability:  unavailable

 

Electronical lecture notes

Availability:  unavailable

 

Literature

Recommended literature
Mahmoud Hassaballah and Ali Ismail Awad. "Deep Learning in Computer Vision Principles and Applications". CRC Press, 2020. isbn: 978-1-138-54442-0
Ian Goodfellow, Yoshua Bengio, and Aaron Courville. "Deep Learning". MIT Press, 2016. ISBN: 978-0-262-03561-3
Mohamed Elgendy. "Deep Learning for Vision Systems". Manning Publications, 2020. ISBN: 978-1-61729-619-2

Assigned Degree Programs


This module is used in the following Degree Programs (new System):

Studiengang / StuPOStuPOsVerwendungenErste VerwendungLetzte Verwendung
Computer Engineering (M. Sc.)115SoSe 2023SoSe 2024
Computer Science (Informatik) (M. Sc.)112SoSe 2023SoSe 2024
Elektrotechnik (M. Sc.)16SoSe 2023SoSe 2024

Miscellaneous

No information