Workpackage

Bringing multimodal environmental perception in embedded AI systems to a performance that enables a wide application in the real world
Bringing multimodal e...
Adaptation of deep learning architectures to the specific characteristics of multimodal environmental perception
Adaptation of deep...
Reduction of the amount of training data that needs to be collected
Reduction of the a...
Development of optimized dedicated Kl implementations for an energy and privacy promoting edge computing use
Development of opt...
TP1: Lukowicz
Physical constraints, pins and inductive bias
TP1: Lukowicz...
TP2: Dengel
semantic information for data- and energy-efficient recognition
TP2: Dengel...
TP3: Stricker
Continuous Learning und SNNs
TP3: Stricker...
TP4: Berns
Data generation by simulation
TP4: Berns...
TP5: Wehn/
Schöbel
Design methodology for dedicated hardware
TP5: Wehn/...
TP6: Ruszkowski/ Plociennik
Smart Factory
Joint experiments and applications, research on semantic school of administration, energy management
TP6: Ruszkowski/ Plo...
TP7: Dörr
Smart Farming
Joint experiments and applications, research on explainability
TP7: Dörr...
MilestoneM6b
independent, superordinate methodology for
the field of multimodal environmental perception
MilestoneM6b...
MilestoneM6c
widely usable tools and datasets
MilestoneM6c...
Desired achievement
Desired achievement
Methodological
top level goals
Methodological...
Specific individual
research work
Specific individual...
Joint evaluation
and demonstration
Joint evaluation...
Tangible results
Tangible resultsText is not SVG - cannot display