Unit
The Visual Intelligence for Transportation (VITA) laboratory at EPFL focuses on developing socially-aware AI systems for transportation and mobility applications. Their research aims to enable self-driving vehicles and delivery robots to coexist safely and efficiently with humans in crowded social scenes. By combining Computer Vision, Machine Learning, and Robotics, the lab works on understanding human behavior, predicting actions, and planning interactions in real-time. Projects include pedestrian behavior prediction, human pose estimation, and crowd-robot interaction. The lab also explores generative models and motion representations for robust and socially-compliant mobility solutions.