Reproduce “Provisioning Robust and Efficient LiDAR Perception for Autonomous Driving” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Tingxu Han, Zijie Shen, An Guo, Lyuke Wang, Chunrong Fang, Zhenyu Chen, Wei Song
The authors provision LiDAR 3D object detectors offline to improve corruption performance while reducing inference costs. They order clean and synthetically corrupted training pairs by how much an ensemble’s predictions differ, and reduce model widths and sampling workloads using data statistics. Across 12 dataset–detector–class combinations, corruption AP gains average 4.03 points, but one declines; tests cover synthetic corruptions only.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.