Perception of Humans
Multi-modal perception of an agent's environment is crucial for interacting naturally with human partners in non-laboratory settings. This applies to robots as well as to intelligent smart-homes or interactive virtual agents. We use multi-modal perception of activities on a global and local human-centered level.
Global activity analysis
Global activity analysis aims at detecting and recognizing interaction partners. We detect humans in the environment using multi-modal cues such as speech, rhythm, timing, attention, positioning, and posture, and person anchoring. Basal 3D information e.g. , is represented in an articulated scene model to keep track of interaction partners and as a prior to identify manipulatable objects .
Identification of interaction relevant cues
In order to interpret interaction situations, we focus on the recognition of semantic events of human interaction partners. With the help of human cues, we analyze joint spaces between humans and robots to research spatial behaviors and enable a smooth transition between far away and face-to-face human-robot interaction .
We also examine human behaviors in specific situations such as at a receptionist desk or while tooth-brushing . Within these situations, we for example detect deviations from standard behavior to support the mentally inhibited.
- SFB 673: Alignment in Communication
- Cluster of Excellence Cognitive Interaction Technology (CITEC)
- Research Institute for Cognition and Robotics (CoR-Lab)
2010 | Journal Article | PUB-ID: 2034623Using Articulated Scene Models for Dynamic 3D Scene Analysis in Vista SpacesPUB | PDF | DOI | arXiv
Beuter N, Swadzba A, Kummert F, Wachsmuth S (2010)
3D Research 1(3): 1-13.
2011 | Journal Article | PUB-ID: 2481769How Can I Help? - Spatial Attention Strategies for a Receptionist RobotPUB | PDF | DOI | WoS
Holthaus P, Pitsch K, Wachsmuth S (2011)
International Journal of Social Robotics 3(4): 383-393.
2012 | Conference Paper | PUB-ID: 2423452User Behavior Recognition For An Automatic Prompting System - A Structured Approach based on Task AnalysisPUB | DOI
Peters C, Hermann T, Wachsmuth S (2012)
In: Proceedings of the 1st Int. Conf. on Pattern Recognition Applications and Methods (ICPRAM)., 2. SciTePress: 162-171.
Recent Best Paper/Poster Awards
Philippsen A, Reinhart F, Wrede B (2016)
International Conference on Development and Learning and on Epigenetic Robotics (ICDL-EpiRob)
Richter V, Carlmeyer B, Lier F, Meyer zu Borgsen S, Kummert F, Wachsmuth S, Wrede B (2016)
International Conference on Human-agent Interaction (HAI)
Carlmeyer B, Schlangen D, Wrede B (2016)
International Conference on Human Agent Interaction (HAI)