É«É«À²

Technological advancements have enabled autonomous vehicles (AVs) to operate effectively in controlled environments. However, deploying these vehicles in urban settings remains a significant challenge due to the complexities of interacting and communicating with pedestrians. Addressing this limitation in autonomous driving technology is crucial for enhancing the understanding and interaction with human road users. Our research tackles this issue by developing computational models that accurately simulate human behavior and their interactions with autonomous vehicles. Our work in this area is done in collaboration with Finnish Center for Artificial Intelligence (FCAI).

What we do:

• Jointly learning trajectory and intent prediction for modeling human behavior
• Virtual Reality-based scenario generation to model human behavior and interactions with AVs
• Incorporating foundation models to understand the AV-human interactions
• Building courteous AVs, that understand and communicate with humans and are socially-aware about environments

Keywords:

Large and Visual Language Models; Human-Machine Interaction; Perception for Autonomous Driving; Autonomous Driving