General Informationen for the Course "Context Recognition Architectures" (5803V and 5803UE)

Course Type: Mandatory course
Holding: in the summer term
Course Number: Lecture: 5803V and Tutorial: 5803UE
Hours per Week: 3 (2 Lecture, 1 Tutorial, 0 Laboratory)
ECTS-Credits: 5,0
Dates Lecture: Mon., 14:00 - 16:00, (JUR) HS 14
Dates Tutorial: Wed., 10:00 - 11:00, (ITZ) SR 002
Expected Number of Participants: ca. 30
Begin of the Lecture: 24.04.2017
Begin of the Tutorial: 26.04.2017

Course Description

Content:  
The Spirit of Context Aware Computing: Historical Background, Pioneering and Influential Work, Application Cases, Current Research Trends, Outlook
Sensors: Overview of available sensors especially suited for the use in context recognition Architectures. (e.g., inertial measurement units for wearable activity recognition; Environmental Sensors, Biosignal, Smartphone as a sensing Platform, etc.). Sensor Node Design & Communication (Bluetooth, ZigBee, etc.)
Context Aware System Design Principles I: Introducing the Activity and Context recognition chain to transform raw sensor data towards semantic information.
Context Aware System Design Principles II: Detailed walkthrough and methodological explanation of the necessary steps in the Activity and Context recognition chain.
Context Aware System Design Principles III: Combination of learned methodologies towards a realtime, activity and context recognition architecture. Identification of specific shortcommings of bottom-up vs. top-down architectures and their possible solution.
Identification, Presence & Tracking: Identifying Human & Things; Technologies for Identification (RFID, NFC), Positioning, Orientation, Smart Dust, Surfacewave Transponder, Visual Codes (QR), Artificial Noses, Selected Application Cases (e.g., Driver Identity-/Activity Recognition)
Social Aware Systems & Patterns: Social Computing in general, Algorithms based on graph theory for community detection,
  SmartPhone as a Sensing Platform on multiple scale and for Crowd Context Detection
Looking into the future: Introduction into Time Series Prediction, Multiplicative Time Series Model, ARMA, ARIMA, Context Prediction based on State Space Models (HMM)
Security Matters?: Security and Privacy Definitions, Solove’s Privacy Taxanomy, Legal Issues, Fair information principles, UbiComp Implications, Shamir Tags, Critical Examples of RFID and Smart Devices.
Wearable Computing: Off the shelf technology review and application scenarios discussion (Glasses, SmartWatches, FitnessTrackers).
Energy Efficient Design Methodologies: Design specifics to ensure low power consumption (in terms of soft- and hardware techniques) resulting in long (and/or optimized )
   
Previous knowledge required: none
Previous knowledge expected: Systementwicklungsprojekt (SEP) oder MES Praktikum, Programmierung in Java oder Programmierung 1 und Programmierung 2, Einführung in die Kontexterkennung
Objective (expected results of study and acquired competences):  
Skills: Students know the basic design principles, representative and importent projects in the area oft the course, and evaluation criteria’s for Context Recognition Architectures. Students know how recognition methodologies are implemented in real time systems by applying state of the art machine learning and pattern classification methodologies. Students know the fundamental theoretic and practical problems when designing context recognition architectures.
Abilities: Students can use their theoretical knowledge about the single steps of the so called “Activity Recognition Chain” ( i.e. sensor selection, sensor sampling, segmentation, feature extraction, classification, fusion, and symbolic processing/reasoning) to apply, discuss, and implement it.
Competence: Students learn theoretical and practical competencies (i) in the conception, (ii) in the design, (iii) in the implementation and (iv) in the evaluation of Context Recognition Architectures. During the practical tasks, special focus is put on the reusability of the developed software components to make them easily (re-)usable in future application scenarios.
Languages of instruction: English
Learning Organisation: Präsentation mit Projektor, Gruppenarbeit, Wiki
Scheduled dates: see StudIP
Course criteria & registration: Please refer to the official announcements of the university office (Studiensekretariat). They will be announced usually 4-6 weeks after the start of the term. Please direct all information requests directly and solely to the university office (Studiensekretariat).

Exam Information

Assessment (exam method and evaluation): 90-minütige Klausur oder mündliche Prüfung (ca. 20 Minuten) oder Portfolio-Prüfung;
   
Portfoliobestandteile sind: - Schriftliche Teilprüfung (66%)
  - Praktischer Teil (33%)
  o Systemimplementierung
  o Dokumentierter und funktionsfähiger Quelltext
  o Laufende, fortzuschreibende technische Teilberichte zur Zusammenfassung zu einem Gesamtdokument.
  o Abschlusspräsentation
   
  Die Bearbeitung der Portfolio-Leistungen erfolgt begleitend zur Lehrveranstaltung. Die Bearbeitungszeit der einzelnen Bestandteile der Portfolioprüfung darf dabei 3 Wochen nicht übersteigen. Die letzte Leistung ist bis spätestens 4 Wochen nach Ende der Vorlesungszeit zu erbringen.
  Der Umfang der Abschlusspräsentation soll dabei 15 Minuten umfassen und durch geeignete Medien und Präsentationsformen unterstützt werden.
  Die genauen Anforderungen werden vom Dozierenden zu Beginn der Veranstaltung bekanntgegeben.
  Die Modulnote entspricht der Note der Prüfung.
   
Exam dates & registration: will be announced.
Number of exam dates per semester: 1

Further Information:

StudIP: https://studip.uni-passau.de/studip/dispatch.php/course/overview?cid=24ad6baa5b11e581568257fc714ba0cc
Wiki: -
Moodle: -
Recommended Reading: Wird vom Dozent/ Dozent / von der Dozentin bekannt gegeben gegeben.
  Die Literatur wird in Abhängigkeit der konkreten Aufgabenstellung ausgewählt und bekanntgegeben.