Probabilistic models of cognition
- hierarchical multimodal model of cognition (vision + language)
- probabilistic model of language acquisition
- iCub robot experiments (cooperation with prof. Angelo Cangelosi, CRNS lab, Plymouth) – grounding language in vision (implementation of the hierarchical model of language acquisition into the robotic platform)
- cooperation with Nicolaos Mavridis (Ripley robot extension)
- symbol grounding
- cooperation with Angelo Cangelosi (implementation of the architecture in iCub robot platform), see youtube video for iCub language acquisition
- PhD thesis: Probabilistic hierarchical model of language acquisition
- GACR project MIRRACLE: Multimodal representation of robotic actions and tasks applied in learning by demonstration
Imitation learning with a support of language
- multimodal integration (linguistic and visual information about the performed task)
- 3D semantic segmentation of objects
- trajectory tracking
- activity recognition
- specifying tasks by NL and demonstration, see YouTube video
- TACR project: Imitation learning of industrial robots with a support of language
- CROW project: Collaborative robotic workspace of a future
Robot self-calibration
- robot self-calibration using multimodal information (tactile and visual feedback), see YouTube video,
- automatic calibration of both humanoid and industrial robots via calibration toolbox: see related publications
- optimization of parameters using non-linear least squares, pose selection
- GACR: Automatic calibration of robots and safe interaction with human inspired by the primate brains
EEG signal processing and analysis
- focus on higher cognitive functions
- mentally gifted children vs. normal population