|
|
Einrichtungen >> Technische Fakultät (TF) >> Department Artificial Intelligence in Biomedical Engineering (AIBE) >>
|
Lehrstuhl für Maschinelles Lernen und Datenanalytik
| |
Bewegungsanalyse und biomechanische Grenzgebiete [BABG] -
- Dozent/in:
- Anne Koelewijn
- Angaben:
- Vorlesung, 2 SWS, ECTS: 2,5, für FAU Scientia Gaststudierende zugelassen
- Termine:
- Zeit/Ort n.V.
- Studienrichtungen / Studienfächer:
- WF MT-BA ab 4
WF MT-MA ab 1
WF MT-MA-BDV ab 1
WF MT-BA-BV ab 4
WF MT-MA-MEL ab 1
WF MT-MA-IDP ab 1
WF MT-BA-GP ab 4
- Voraussetzungen / Organisatorisches:
- Für diese Lehrveranstaltung ist eine Anmeldung erforderlich.
Die Anmeldung erfolgt über: https://www.vhb.org/
- Inhalt:
- • Anatomie des menschlichen Bewegungsapparates
Muskeln, Sehnen, Bänder, Knochen, Knorpel• Gelenkmechanik • Kinematik
Bewegungsanalyse und Motion-Capturing-Systeme • Kinetik
Kraft- & Druckmessplatten, Bodenreaktionskräfte • Elektromyographie • 3D-Modellierung in der Biomechanik Segmentierung, 3D-Modelle • Simulation
FEM, MKS
- Empfohlene Literatur:
- Relevante Literatur ist im online-Kurs zu den jeweiligen Kapiteln angegeben.
|
| |
Catching your eyes: AI-driven modeling and analysis of eye-tracking data [ETS] -
- Dozent/in:
- Dario Zanca
- Angaben:
- Seminar, 2 SWS, ECTS: 2,5, für FAU Scientia Gaststudierende zugelassen
- Termine:
- Di, 12:15 - 13:45, 00.010
Organisation and slides via StudOn.
- Voraussetzungen / Organisatorisches:
- Assignment: mailto: dario.zanca@fau.de
The grade is based on a presentation and a report, both counting 50% of the final grade.
- Inhalt:
- Learning objectives:
Be familiar with an eye-tracking experimental setup and eye-tracking data.
Knowledge of the common eye-tracking data analysis techniques.
Knowledge of the state-of-the-art saliency and scanpath models to predict human visual attention.
Contents
Seeing is a complex activity. Humans perform eye movements to actively seek for useful information, while regulating pupil size to control the amount of light to be captured. Eye-tracking can be used to record the eye’s activity. It is a powerful tool to study human gaze behavior and it can be used to assess the health condition of individuals.
The aim of this seminar is to become familiar with eye-tracking data and their use in different domains, from neuroscience and artificial intelligence (to understand and simulate human attention), to medicine and psychology (to identify eye-tracking based biomarkers).
Different methods will be introduced and compared. Students will study on state-of-the-art papers and present the details of the chosen topic described in the papers. Alternatively, the student may work on experimental task and present the result of applying state of the art methods.
- Empfohlene Literatur:
- • Itti, L., Koch, C., & Niebur, E. (1998). A model of saliency-based visual attention for rapid scene analysis. IEEE Transactions on pattern analysis and machine intelligence, 20(11), 1254-1259.
• Borji, A., & Itti, L. (2012). State-of-the-art in visual attention modeling. IEEE transactions on pattern analysis and machine intelligence, 35(1), 185-207.
• Judd, T., Ehinger, K., Durand, F., & Torralba, A. (2009, September). Learning to predict where humans look. In 2009 IEEE 12th international conference on computer vision (pp. 2106-2113). IEEE.
• Zanca, D., & Gori, M. (2017, December). Variational laws of visual attention for dynamic scenes. In Proceedings of the 31st International Conference on Neural Information Processing Systems (pp. 3826-3835).
• Zanca, D., Melacci, S., & Gori, M. (2019). Gravitational laws of focus of attention. IEEE transactions on pattern analysis and machine intelligence, 42(12), 2983-2995.
• Zanca, D., Gori, M., Melacci, S., & Rufa, A. (2020). Gravitational models explain shifts on human visual attention. Scientific Reports, 10(1), 1-9.
• Bellet, M. E., Bellet, J., Nienborg, H., Hafed, Z. M., & Berens, P. (2019). Human-level saccade detection performance using deep neural networks. Journal of neurophysiology, 121(2), 646-661.
• Piu, P., Serchi, V., Rosini, F., & Rufa, A. (2019). A cross-recurrence analysis of the pupil size fluctuations in steady scotopic conditions. Frontiers in neuroscience, 13, 407.
• Zénon, A. (2017). Time-domain analysis for extracting fast-paced pupil responses. Scientific reports, 7(1), 1-10.
• Bargagli, A., Fontanelli, E., Zanca, D., Castelli, I., Rosini, F., Maddii, S., ... & Rufa, A. (2020). Neurophthalmologic and Orthoptic Ambulatory Assessments Reveal Ocular and Visual Changes in Patients With Early Alzheimer and Parkinson's Disease. Frontiers in Neurology, 11.
- Schlagwörter:
- eye-tracking, human visual attention, biomarkers, artificial intelligence
|
| |
Green AI - AI for Sustainability and Sustainability of AI [GREENAI] -
- Dozentinnen/Dozenten:
- Eva Dorschky, René Raab, Björn Eskofier
- Angaben:
- Seminar, 2 SWS, benoteter Schein, ECTS: 5, für FAU Scientia Gaststudierende zugelassen, There are no more free places in the SS 2022.
- Termine:
- Do, 10:15 - 11:45, 00.010
Einzeltermin am 21.7.2022, 10:15 - 11:45, 01.151-128
- Studienrichtungen / Studienfächer:
- WPF INF-MA ab 1
WF MT-MA ab 1
WPF AI-MA ab 1
WPF INF-BA-SEM ab 5
WPF CE-MA-SEM ab 1
|
| |
Human Computer Interaction [HCI] -
- Dozent/in:
- Björn Eskofier
- Angaben:
- Vorlesung, 3 SWS, ECTS: 3,75
- Termine:
- Di, Do, 8:15 - 9:45, H10
Die erste Veranstaltung findet am 28.04 um 08:15 in H10 statt.
- Studienrichtungen / Studienfächer:
- WPF INF-MA ab 1
WPF INF-BA ab 5
WF CE-MA ab 1
WF ICT-MA ab 1
WF IuK-BA ab 5
WPF ICT-MA-ES ab 1
WPF ICT-MA-MPS ab 1
WPF ICT-MA-NDC ab 1
WF EEI-MA ab 1
WF EEI-BA ab 5
WPF MT-MA-BDV ab 1
WF MT-MA-GPP ab 1
WF MT-MA-MEL ab 1
WPF MT-BA ab 5
WPF ASC-MA ab 1
WPF AI-MA ab 1
- Voraussetzungen / Organisatorisches:
- Folien zur Vorlesung und Organisation über Studon.
- Inhalt:
- Studon Kurs:
https://www.studon.fau.de/studon/goto.php?target=crs_4380069
- Schlagwörter:
- human-computer interaction, Mensch-Maschine-Schnittstelle, grafische Benutzerschnittstellen, mobile Mensch-Computer-Interaktion, Mensch-Maschine-Interaktion im Fahrzeug, ubiquitäre und eingebettete interaktive Systeme
|
| |
Human Computer Interaction Exercises [HCI-E] -
- Dozent/in:
- Madeleine Flaucher
- Angaben:
- Übung, 1 SWS, ECTS: 1,25
- Termine:
- Di, 12:15 - 13:45, H3 Egerlandstr.3
- Studienrichtungen / Studienfächer:
- WPF INF-MA ab 1
WF INF-BA ab 5
WF CE-MA ab 1
WF ICT-MA ab 1
WF IuK-BA ab 5
WPF ICT-MA-MPS ab 1
WPF ICT-MA-ES ab 1
WPF ICT-MA-NDC ab 1
WF EEI-MA ab 1
WF EEI-BA ab 5
WPF MT-MA-BDV ab 1
WPF MT-BA ab 5
WPF ASC-MA ab 1
WPF AI-MA ab 1
- Schlagwörter:
- human-computer interaction, Mensch-Maschine-Schnittstelle, grafische Benutzerschnittstellen, mobile Mensch-Computer-Interaktion, Mensch-Maschine-Interaktion im Fahrzeug, ubiquitäre und eingebettete interaktive Systeme
|
| |
Innovationslabor für Wearable und Ubiquitous Computing [InnoLab] -
- Dozentinnen/Dozenten:
- Björn Eskofier, Matthias Zürl, Michael Nissen, Marlies Nitschke, Nils Roth, Johannes Link, Mohamad Wehbi, Imrana Abdullahi Yari, Alzhraa Ahmed, Ann-Kristin Seifer, Misha Sadeghi
- Angaben:
- Praktikum, 4 SWS, benoteter Schein, geeignet als Schlüsselqualifikation, für FAU Scientia Gaststudierende zugelassen
- Termine:
- Di, 16:15 - 17:45, 00.010
Do, 12:15 - 13:45, 00.010
- Studienrichtungen / Studienfächer:
- WF INF-MA ab 1
WF MT-MA ab 1
WF MB-MA ab 1
WF ME-MA ab 1
WF WING-MA ab 1
WF CE-MA ab 1
WF IIS-MA ab 1
WF Ph-MA ab 1
WPF DS-MA ab 1
- Voraussetzungen / Organisatorisches:
- Themenvergabe und Terminfindung in der ersten Woche des Semesters. Die Vergabe der Plätze im Kurs erfolgt nach einem Aufnahmeantrag. In diesem sollen Studierende ihren Studiengang, ihr Fachsemester und den angestrebten Abschluss nennen (Bachelor/Master). Sind zu viele Anmeldungen eingegangen und der Kurs ist voll, gibt es eine Warteliste.
Das Praktikum steht Studierenden der genannten Studienrichtungen ab dem 5. Bachelorsemester (und aller Mastersemester) ebenfalls offen. Für weitere Studiengänge und ECTS-Verteilungen bitte unter Matthias.Zuerl@fau.de nachfragen.
Weitere Informationen zu der Lehrveranstaltung können auch unserer Webseite entnommen werden: https://www.mad.tf.fau.de/teaching/innolab/ ,
Anmeldung: https://www.studon.fau.de/crs4362400.html
Die Anmeldung ist möglich ab dem 14.03.2022 bis einschließlich 15.04.2022
- Inhalt:
- Mini-Computer, die unseren Lebensrhythmus dokumentieren, EKG-Sensoren, die jedes Detail aufzeichnen, Brillen, die uns in eine andere Realität versetzen – diesen Technologien begegnen wir mittlerweile ständig im Alltag. Im Innovationslabor für Wearable und Ubiquitous Computing werden solche Technologien von Studierenden entwickelt und gleichzeitig aufgezeigt, wie man mit diesen ein eigenes Startup gründen könnte.
Die innovativen Technologien werden dabei prototypisch in Gruppenarbeit (5-8 Studierende) unter Nutzung von agilen Entwicklungsmethoden (Scrum) geschaffen. Den Studierenden steht dabei der Zugang zum Innovationslabor offen, welches mit der nötigen Infrastruktur für die Entwicklung der Prototypen ausgestattet ist. Die Ideen für die Projekte stammen dabei entweder von kooperierenden Firmen oder von den Studierenden selbst.
Neben dem Prototyping erlernen die Teilnehmer in Tutorials die Grundlagen für innovatives Arbeiten wie Design Thinking und Patentrecherche. Zudem wird ihnen beigebracht, wie sie nach der Entwicklung ihre Ideen schützen und gegebenenfalls an den Markt bringen können.Lernziele und Kompetenzen:
Ideation, Design Thinking
Patentrecherche, Marktanalysetechniken
Agile Entwicklungsmethoden (Scrum)
Prototyping
Sicherung geistigen Eigentums
Einführung in Entrepreneurship, Startup Finanzierung
. Prüfungsleistung
Der Umfang der Leistung im Innovation Lab setzt sich aus vier Teilbereichen zusammen (in Klammern ist der Anteil an der Gesamtnote angegeben):
Teampräsentation - 30 min (30%)
Abschlussvortrag - 10 min (10 %)
Hardware/Software Development, Scrum Meetings, Practical work (40%)
Abschlussdokumentation/Report - ca. 3 - 6 Seiten pro Studierendem (20 %)
Nach dem erfolgreichen Absolvieren erhalten die Studierenden 10 ECTS.
|
| |
Legged Locomotion of Robots [LLR] -
- Dozent/in:
- Anne Koelewijn
- Angaben:
- Seminar, 2 SWS, benoteter Schein, ECTS: 2,5, für FAU Scientia Gaststudierende zugelassen
- Termine:
- Di, 10:15 - 11:45, 00.010
- Studienrichtungen / Studienfächer:
- WF EEI-BA ab 6
WF EEI-MA ab 1
WPF INF-MA ab 1
WF MT-MA-BDV ab 1
WF MT-MA-GPP ab 1
WF MT-MA-MEL ab 1
- Schlagwörter:
- Robotics, Humanoids, Legged Locomotion, Walking Robots, Control, Dynamic Walking, Stability, Energetics
|
| |
Maschinelles Lernen und Datenanalytik für Industrie 4.0 [MADI40] -
- Dozentinnen/Dozenten:
- Björn Eskofier, Johannes Roider, Christoph Scholl, Lukas Schmidt
- Angaben:
- Seminar, 2 SWS, benoteter Schein, ECTS: 5, nur Fachstudium, für FAU Scientia Gaststudierende zugelassen, Registration via mail to johannes.roider@fau.de
- Termine:
- Mi, 16:15 - 18:00, 00.010
Starts April 27th 2022
- Studienrichtungen / Studienfächer:
- WPF MT-MA-BDV ab 1
WPF INF-MA ab 1
WPF CE-MA ab 1
WF ASC-MA ab 1
WF ICT-MA ab 1
WPF AI-MA ab 1
- Voraussetzungen / Organisatorisches:
- Registration via e-mail to johannes.roider@fau.de Registration period: 25.02.-04.05.2022
The seminar will be held face-to-face.
Requirements:
Prior knowledge of machine learning via courses like PA, IntroPR, PR, DL, MLTS, CVP or equivalent (ideally first project experiences) is expected! Motivation to explore scientific findings (e.g. via literature research) Motivation to code and analyze data
Please state your previous experience in machine learning (e. g. Which courses did you take? Which project experience do you have?) when registering for the course. Examination:
50% of grade: Presentation + demo (20 minutes)
50% of grade 4 pages IEEE standard paper (excluding references) (+ code submission)
Attendance of all meetings is required.
- Inhalt:
- Contents
Companies in all kinds of industries are producing and collecting rapidly more and more data from various sources. This is enabled by technologies such as the Internet of Things (IoT), Cyber-physical system (CPS) and cloud computing. Hence there is an increasing demand in industry and research for students and graduates with machine learning and data analytics skills in the Industry 4.0 context.
In this Seminar the Industry 4.0 term will include adjacent fields like the medical device or the automotive sector. Aim of this seminar is to give students insights about state-of-the-art machine learning and data analytics methods and applications in Industry 4.0 and adjacent fields. Students will mainly work independently on either a implementation centric or a research centric topic. The implementation centric topics will focus primarily on the implementation of algorithms and analytical components, while the research centric topic will focus on researching and structuring literature on a specific field of interest. Several potential topics will be provided but students are also encouraged to propose their own topics (please discuss with teaching staff beforehand). Topics covered will include but are not limited to:
Best practices for presentation and scientific work
Brief overview of current hot topics in the field of machine learning and data analytics for Industry 4.0 (e.g. deep learning for predictive maintenance and process mining for usage analysis)
Data acquisition (what kind of data can be acquired? Identification of publicly available data sets) and storage (how can data be stored efficiently?)
Machine learning and data analytics methodologies (Support vector machines, Hidden Markov models, Deep learning, Process mining, etc.) for industrial data (sensor data, event logs, ...)
The seminar will include talks by corresponding lecturer and invited experts in the domain. Furthermore, students will present results from literature research and data analytics projects. Learning Objectives and Competencies
Students will develop an understanding of the current hot field of machine learning and data analytics for Industry 4.0 / healthcare / automotive
Students will learn to research and present a topic within the context of machine learning and data analytics for Industry 4.0 / healthcare / automotive independently
Students will learn to identify opportunities, challenges and limitations of corresponding ML approaches for Industry 4.0 / healthcare / automotive
Students will develop the skill to identify and understand relevant literature and to present their finding in a structured manner
Students will learn to present implementation and validation results in form of a demonstration and/or report
- Empfohlene Literatur:
- Lei, Yaguo, Naipeng Li, Liang Guo, Ningbo Li, Tao Yan, and Jing Lin. “Machinery Health Prognostics: A Systematic Review from Data Acquisition to RUL Prediction.” Mechanical Systems and Signal Processing 104 (May 2018): 799–834.https://doi.org/10.1016/j.ymssp.2017.11.016.
Rojas, Eric, Jorge Munoz-Gama, Marcos Sepúlveda, and Daniel Capurro. “Process Mining in Healthcare: A Literature Review.” Journal of Biomedical Informatics 61 (June 1, 2016): 224–36. https://doi.org/10.1016/j.jbi.2016.04.007. Wil M. P. van der Aalst. „Process Mining: Data Science in Action” 2nd edition, Springer 2016. ISBN 978-3-662-49851-4 Wang, Lihui, and Xi Vincent Wang. Cloud-Based Cyber-Physical Systems in Manufacturing. Cham: Springer International Publishing, 2018. https://doi.org/10.1007/978-3-319-67693-7.
- Schlagwörter:
- Machine Learning, Data Analytics, Process Mining, Predictive Maintenance, Industry 4.0, Healthcare, Automotive
|
| |
Reinforcement Learning [RL] -
- Dozent/in:
- Christopher Mutschler
- Angaben:
- Vorlesung, 2 SWS, benoteter Schein, ECTS: 2,5
- Termine:
- Do, 8:30 - 10:00, Zoom-Meeting
- Studienrichtungen / Studienfächer:
- WF ASC-MA ab 1
WF INF-MA ab 1
WF CE-MA-TA-MT ab 1
WF MT-MA ab 1
WF ME-MA ab 1
WPF ME-BA-MG6 4-6
WPF ME-MA-MG6 1-3
WF CME-MA ab 1
WF ICT-MA ab 1
WPF DS-MA-AI ab 1
- Inhalt:
- Reinforcement Learning (RL) is an area of Machine Learning that has recently made large advances and has been publicly visible by reaching and surpassing human skill levels in games like Go and Starcraft. These successes show that RL has the potential to transform many areas of research and industry by automatizing the development of processes that once needed to be engineered explicitly.
In contrast to other machine learning paradigms, which require the presence of (labeled or unlabeled) data, RL considers an agent that takes actions in an environment and learns from resulting feedback. The agent tries to maximize a reward signal that it receives for desirable outcomes, while at the same time trying to explore the world in which it operates to find yet unknown, potentially more rewarding action sequences–a dilemma known as the exploration-exploitation tradeoff. Recent advances in machine learning based on deep learning have made RL methods particularly powerful since they allow for agents with particularly well performing models of the world. The lecture will start with introductory lectures to RL where we cover the foundations of RL (i.e., Markov decision processes and dynamic programming techniques) before we go to model-free prediction and control algorithms such as TD-learning, SARSA and Q-learning. We will also get the general idea behind value function approximation techniques such as Deep Q-Networks (DQN) and study advanced policy-gradient and actor-critic methods including TRPO and PPO. We will end with focus sessions on advanced topics such as model-based RL, offline RL, explainable RL, and exploration-exploitation.
- Empfohlene Literatur:
- While there is particular literature given in the slides of the videos the following list serves as a general basis to get into the topic but also to go deeper at particular points.
Richard S. Sutton and Andrew G. Barto. 2018. Reinforcement Learning: An Introduction. A Bradford Book, Cambridge, MA, USA. Bellman, R.E. 1957. Dynamic Programming. Princeton University Press, Princeton, NJ. Republished 2003: Dover, ISBN 0-486-42809-5. Csaba Szepesvari and Ronald Brachman and Thomas Dietterich. 2010. Algorithms for Reinforcement Learning. Morgan and Claypool Publishers. Warren B. Powell. 2011. Approximate Dynamic Programming. Wiley. Maxim Lapan. 2020. Deep Reinforcement Learning Hands-On: Apply modern RL methods to practical problems of chatbots, robotics, discrete optimization, web automation, and more, 2nd Edition. Packt Publishing. Dimitri P. Bertsekas. 2017. Dynamic Programming and Optimal Control. Athena Scientific. Miguel Morales. 2020. grokking Deep Reinforcement Learning. Manning. Laura Graesser and Keng Wah Loon. 2019. Foundations of Deep Reinforcement Learning: Theory and Practice in Python. Addison-Wesley Data & Analytics.
|
|
 |
|
|
UnivIS ist ein Produkt der Config eG, Buckenhof |
|
|