Faculty of Informatics / Mathematics

Department of Artificial Intelligence / Cognitive Robotics

three_kings
HTW Dresden/Roth

Background

The research activities of the Artificial Intelligence Lab are focused on the development and evaluation of mobile, interactive service and assistance systems. Application scenarios of such systems are for example

  • Guidance and information systems in public buildings (museums, airports, train stations, exhibitions, etc.),
  • Assistance systems in stores (do-it-yourself stores, shopping malls),
  • Systems at home, in particular in the field of ambient assisted living,
  • as well as robots in industrial environments (e.g. mobile measure and surveillance systems).

Mobile for us means, that we typically, but not necessarily, work with robotic systems.

Interactive means, that the function of such systems is mainly driven by interaction with their human users.

Fields of interest and methods

The addressed research problems contain on the one hand robust, adaptable robot navigation in every day environments with emphasis on:

  • Environment modeling,
  • Self-localization,
  • SLAM algorithms, as well as
  • Adaptive approaches for path planning and motion control.

On the other hand, for the interaction with people the areas

  • Person perception (detection and tracking),
  • Analysis of motion patterns (gestures, movement trajectories),
  • Detail analysis of faces (identification; age, gender, and emotion estimation),
  • Speech recognition, as well as
  • Concepts for an adaptive dialogue control are of particular interest.

Characteristic for all mentioned areas is their multi-modality. This comes with

  • the usage of several sensor channels for perceiving the environment
    • 2D/3D cameras,
    • Sonar and laser scanners,
    • Touchscreens,
    • Microphone,
  • as well as different ways for the system to articulate itself
    • Speech,
    • Gaze control,
    • Graphics and Animations,
    • Projections.

The applied spectrum of methods shows a wide variety and ranges from 

  • Modern approaches of neuro informatics,
  • Fuzzy logic and neuro-fuzzy-concepts, over
  • Evolutionary and genetic algorithms, up to
  • Probabilistic techniques.

 

PD Prof. Dr.-Ing. habil. Hans-Joachim Böhme

Head of department

Response times:

Person