Survey of DL-Based LiDAR Super-Resolution For Autonomous Driving (University College London)

University College London researchers published a comprehensive survey on deep learning-based LiDAR super-resolution for autonomous driving. The paper addresses the challenge of expensive high-resolution LiDAR sensors by using deep learning to enhance sparse point clouds from affordable low-resolution sensors. The survey categorizes existing approaches into CNN-based architectures, model-based deep unrolling, implicit representation methods, and Transformer and Mamba-based approaches. Current trends include range image representation, extreme model compression, and real-time inference prioritization for practical deployment. The paper identifies open challenges and future research directions for advancing LiDAR super-resolution technology.