Intersection Scenario
Intersection scenarios, encompassing diverse applications from traffic management to object detection and 3D vision, focus on improving efficiency and safety where multiple entities converge. Current research utilizes various approaches, including reinforcement learning for traffic control, deep learning models (like UNet and DETR variants) for image processing and object tracking, and novel loss functions (e.g., Gr-IoU, Focaler-IoU) to enhance accuracy. These advancements are significant for improving autonomous systems, optimizing urban infrastructure, and enhancing the reliability of machine learning models in complex environments.
Papers
A Realistic Cyclist Model for SUMO Based on the SimRa Dataset
Ahmet-Serdar Karakaya, Konstantin Köhler, Julian Heinovski, Falko Dressler, David Bermbach
Model-Based Deep Learning: On the Intersection of Deep Learning and Optimization
Nir Shlezinger, Yonina C. Eldar, Stephen P. Boyd
HARL: A Novel Hierachical Adversary Reinforcement Learning for Automoumous Intersection Management
Guanzhou Li, Jianping Wu, Yujing He