Learning Socially Aware Behaviors: Developing Autonomous Systems with Situational Awareness in Dynamic Environments
Bidragets beskrivning
Autonomous systems are becoming part of our everyday lives. For these machines to operate safely and naturally alongside people, they must do more than simply detect objects and avoid collisions. They need to understand human behavior. For example, a car should recognize when a pedestrian intends to cross the street, not just that someone is standing nearby.This project develops a new kind of perception framework called Socially Aware Perception (SA2S), which allows autonomous systems to interpret social cues like gaze, attention, and intent. By combining information from cameras, radar, and Lidar with insights from human gaze and cognitive behavior, the system learns to anticipate what people are likely to do next. It also uses probabilistic reasoning to assess risk and uncertainty, helping machines make safer, more understandable decisions. The outcomes will contribute to safer and more trustworthy autonomous systems that can interact with people in a natural, human-centered way.
Visa merStartår
2026
Slutår
2030
Beviljade finansiering
Finansiär
Finlands Akademi
Typ av finansiering
Akademiforskare
Beslutfattare
Forskningsrådet för naturvetenskap och teknik
09.06.2026
09.06.2026
Övriga uppgifter
Finansieringsbeslutets nummer
377983
Vetenskapsområden
El-, automations- och telekommunikationsteknik, elektronik
Forskningsområden
Automaatio- ja systeemitekniikka
Identifierade teman
artificial intelligence, machine learning