Program Overview

Cognitive Systems Lab

The Cognitive Systems Lab is a research institution that focuses on various aspects of cognitive systems, including biosignals, human-computer interaction, and machine learning.

Research Areas

• BiosignalsLab
• Silent voice communication
  • CSL-EMG_Array Corpus
  • Special Session: Novel Paradigms for Direct Synthesis based on Speech-Related Biosignals
• Cognitive Adaptive Interaction Systems
• Systems for people with dementia
• Brain activity modeling
• Automatic speech recognition
• Activity detection
• sensORder: Artifact classification during biosignal acquisition
• Gesture recognition
• Voice communication from neural data

Teaching

The lab offers various courses, including:

• Theses and more
• Winter semester 2025/26
  • Selected problems of cognitive systems
  • Automatic speech recognition
  • Basics of machine learning
  • Block course: Brain pattern recognition
  • Sensor data processing (SdV)
• Summer semester 2025
  • Advanced machine learning
  • Selected problems of cognitive systems
  • Biosignals and user interfaces
  • Block course: Brain pattern recognition
  • Basics of machine learning
  • Seminar Bremen Big Data Challenge

Course: Sensor Data Processing (SdV)

####Description Sensor data processing (SdV) is a course that focuses on the processing of sensor data. The course covers various topics, including:

• Measuring with sensors (camera, microphone, inertial sensors)
• Recording of image, audio and motion data
• From analog to digital measurement signal
• 1D frequency space, Fourier transform, filter
• 2D filter, frequency space, Fourier transform
• Feature extraction for image, audio, motion
• Direct image features
• Segmentation-driven image processing
• Hough transformation
• Classification algorithms
• Development and evaluation of sensor-based systems
• Application examples

Competence Goals

• Understand common sensors such as camera, microphone and inertial sensor
• Understand signals that these sensors measure
• Recognize and use influence (opportunities) to optimize the environment and the recording process for measuring signals
• Understand the transformation from space/time to frequency space for 1D and 2D signals
• Interpret the frequency space clearly
• Master the basic methods for feature extraction, especially for image, sound and motion data
• Know classification approaches
• Design application of sensor data processing at module level
• Evaluate sensor data processing systems

Lecturers

• Profs. Tanja Schultz and Udo Frese

Schedule

• Mondays: 10-12, weekly (from October 21, 2019), lecture, location: NW2 A0242 (step lecture hall)
• Thursday. 09.02.23 09:00 - 13:00 HS 2010 (large lecture hall)

Target Audience

Computer science certificates are accepted for Inf, DM, SE and all degree programs. Cannot be introduced together with BV1 or ABV (before SS20).