Lane Detection
Lane detection, a crucial component of autonomous driving systems, aims to accurately identify and track lane markings on roads for safe navigation. Current research emphasizes improving robustness against challenging conditions like occlusions, shadows, and varying weather, often employing deep learning models such as transformers and convolutional neural networks, sometimes enhanced with techniques like knowledge distillation or attention mechanisms. These advancements are vital for enhancing the safety and reliability of autonomous vehicles and contribute significantly to the broader field of computer vision by pushing the boundaries of real-time, accurate object recognition in complex, dynamic environments.
Papers
Sparse Laneformer
Ji Liu, Zifeng Zhang, Mingjie Lu, Hongyang Wei, Dong Li, Yile Xie, Jinzhang Peng, Lu Tian, Ashish Sirasao, Emad Barsoum
Homography Guided Temporal Fusion for Road Line and Marking Segmentation
Shan Wang, Chuong Nguyen, Jiawei Liu, Kaihao Zhang, Wenhan Luo, Yanhao Zhang, Sundaram Muthu, Fahira Afzal Maken, Hongdong Li
Graph Attention Network for Lane-Wise and Topology-Invariant Intersection Traffic Simulation
Nooshin Yousefzadeh, Rahul Sengupta, Yashaswi Karnati, Anand Rangarajan, Sanjay Ranka