LiDAR Based Perception

LiDAR-based perception focuses on using Light Detection and Ranging (LiDAR) sensors to create accurate 3D representations of the environment for applications like autonomous driving. Current research emphasizes improving the robustness of LiDAR perception against adverse weather conditions (rain, snow, fog) and adversarial attacks, often employing deep learning models such as PV-RCNN and its variants, along with innovative data augmentation and fusion techniques to enhance performance and generalization across diverse sensor setups. These advancements are crucial for ensuring the safety and reliability of autonomous systems and contribute significantly to the development of more robust and dependable perception capabilities in various fields.